CN116718197B - Track processing method and device, electronic equipment and storage medium - Google Patents
Track processing method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN116718197B CN116718197B CN202310999360.6A CN202310999360A CN116718197B CN 116718197 B CN116718197 B CN 116718197B CN 202310999360 A CN202310999360 A CN 202310999360A CN 116718197 B CN116718197 B CN 116718197B
- Authority
- CN
- China
- Prior art keywords
- target
- track
- existing
- candidate
- prediction
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000003860 storage Methods 0.000 title claims abstract description 29
- 238000003672 processing method Methods 0.000 title claims abstract description 21
- 238000000034 method Methods 0.000 claims abstract description 67
- 238000012545 processing Methods 0.000 claims abstract description 16
- 230000008569 process Effects 0.000 description 16
- 238000010586 diagram Methods 0.000 description 15
- 238000005516 engineering process Methods 0.000 description 11
- 238000004590 computer program Methods 0.000 description 10
- 238000001914 filtration Methods 0.000 description 10
- 230000006870 function Effects 0.000 description 8
- 238000004422 calculation algorithm Methods 0.000 description 7
- 238000004891 communication Methods 0.000 description 7
- 230000004927 fusion Effects 0.000 description 7
- 230000003287 optical effect Effects 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 230000005764 inhibitory process Effects 0.000 description 3
- 101100282615 Drosophila melanogaster Gycalpha99B gene Proteins 0.000 description 2
- 230000001133 acceleration Effects 0.000 description 2
- 230000009471 action Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000008034 disappearance Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005265 energy consumption Methods 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 239000004973 liquid crystal related substance Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000000644 propagated effect Effects 0.000 description 2
- 239000004065 semiconductor Substances 0.000 description 2
- 230000001629 suppression Effects 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000005856 abnormality Effects 0.000 description 1
- 238000009825 accumulation Methods 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000003776 cleavage reaction Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000007499 fusion processing Methods 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000007017 scission Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000008685 targeting Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000001131 transforming effect Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
- G01C21/20—Instruments for performing navigational calculations
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/02—Systems using reflection of radio waves, e.g. primary radar systems; Analogous systems
- G01S13/06—Systems determining position data of a target
- G01S13/42—Simultaneous measurement of distance and other co-ordinates
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/02—Systems using reflection of radio waves, e.g. primary radar systems; Analogous systems
- G01S13/50—Systems of measurement based on relative movement of target
- G01S13/58—Velocity or trajectory determination systems; Sense-of-movement determination systems
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/86—Combinations of radar systems with non-radar systems, e.g. sonar, direction finder
- G01S13/862—Combination of radar systems with sonar systems
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/86—Combinations of radar systems with non-radar systems, e.g. sonar, direction finder
- G01S13/865—Combination of radar systems with lidar systems
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/86—Combinations of radar systems with non-radar systems, e.g. sonar, direction finder
- G01S13/867—Combination of radar systems with cameras
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Networks & Wireless Communication (AREA)
- Automation & Control Theory (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The application discloses a track processing method, a track processing device, electronic equipment and a storage medium. The embodiment of the application relates to the technical fields of Internet of vehicles and the like, and can be applied to an intelligent traffic system or an intelligent vehicle-road cooperation system. The method comprises the following steps: performing association matching on a plurality of existing tracks and a plurality of first targets to obtain track association matching results; predicting according to the existing track of the target, and determining a prediction result corresponding to the second target at a prediction time point and a prediction confidence coefficient of the prediction result; and if the prediction confidence coefficient is greater than or equal to the confidence coefficient threshold value, carrying out track association on the predicted position point corresponding to the predicted time point of the second target and the existing track of the target. In the application, the prediction confidence is larger than or equal to the confidence threshold, which indicates that the prediction result of the second target at the prediction time point is accurate, so that the accuracy of the existing track of the second target obtained by carrying out track association on the prediction position point corresponding to the second target at the prediction time point and the existing track of the target is higher.
Description
Technical Field
The present application relates to the field of computer technologies, and in particular, to a track processing method, apparatus, electronic device, and storage medium.
Background
The trajectory (also called trajectory data) refers to data information formed by sampling the course of motion of one or more moving objects. For example, road networks are examples, and the moving object is generally referred to as a vehicle, a pedestrian, or the like. With the rapid development of network technology, the collection of track data and the processing of the collected track data, namely the analysis of the track data by a track modeling method, have become one of the important means for realizing the intensive research of the track data and mining the motion mode of a mobile object in the real world.
In the related art, an observation target observed at an observation time point can be obtained by observing a sensor pair, an existing track of the observation time point is obtained, and the observation target and the existing track are associated to obtain a new existing track. If the existing track at a certain time point is not associated with the observed target, the existing track can be predicted to obtain a predicted position point of the existing target associated with the existing track at the observed time point, and then the association between the existing track and the predicted position point is established to obtain the existing track of the existing target at the time point.
However, the accuracy of predicting the target prediction using the existing method is low, resulting in a low accuracy of the existing trajectory constructed from the predicted location points.
Disclosure of Invention
In view of the above, the embodiments of the present application provide a track processing method, a track processing device, an electronic device, and a storage medium.
In a first aspect, an embodiment of the present application provides a track processing method, where the method includes: performing association matching on a plurality of existing tracks before a target time point and a plurality of first targets observed at the target time point to obtain a track association matching result; if the existing tracks of the targets which are not associated with the first target exist in the plurality of existing tracks according to the track association matching result, predicting according to the existing tracks of the targets, and determining a prediction result corresponding to a second target associated with the existing track of the target before the target time at a prediction time point, wherein the prediction result comprises a prediction position point and a prediction course angle corresponding to the second target at the prediction time point; determining the prediction confidence of the prediction result according to the predicted course angle of the second target corresponding to the predicted time point, the course angle corresponding to the last position point in the existing track of the target and the number of the target predicted position points in the existing track of the target; the target predicted position point is a predicted position point positioned after the last observation position point in the existing track of the target; and if the prediction confidence coefficient is greater than or equal to the confidence coefficient threshold value, carrying out track association on the predicted position point corresponding to the predicted time point of the second target and the existing track of the target.
In a second aspect, an embodiment of the present application provides a track processing apparatus, including: the matching module is used for carrying out association matching on a plurality of existing tracks before the target time point and a plurality of first targets observed at the target time point to obtain track association matching results; the first determining module is used for determining a predicted result corresponding to a second target associated with the target existing track before the target time at a predicted time point according to the target existing track if the target existing track which is not associated with the first target exists in the plurality of existing tracks according to the track association matching result, wherein the predicted result comprises a predicted position point and a predicted course angle corresponding to the second target at the predicted time point; the second determining module is used for determining the prediction confidence of the prediction result according to the predicted course angle corresponding to the second target at the prediction time point, the course angle corresponding to the last position point in the existing target track and the number of the target predicted position points in the existing target track; the target predicted position point is a predicted position point positioned after the last observation position point in the existing track of the target; and the association module is used for carrying out track association on the predicted position point corresponding to the predicted time point of the second target and the existing track of the target if the predicted confidence coefficient is greater than or equal to the confidence coefficient threshold value.
Optionally, the second determining module is further configured to calculate a difference between a predicted heading angle corresponding to the predicted time point of the second target and a heading angle corresponding to the last position point in the existing track of the target; and determining the prediction confidence of the prediction result according to the number and the difference of the target prediction position points.
Optionally, the plurality of first targets are observed by a plurality of sensors; the first determining module is further configured to determine, if there are intersections between the fields of view of the plurality of sensors at the target time point, and a target existing track that is not associated with the first target exists in the plurality of existing tracks according to the track association matching result, predict according to the target existing track, and determine a prediction result corresponding to a prediction time point of a second target associated with the target existing track before the target time point after the target time point.
Optionally, the plurality of first targets are observed by a plurality of sensors; the track association matching result comprises a first association result; the matching module is further used for determining candidate tracks from the plurality of existing tracks and candidate targets from the plurality of first targets according to the observation identifications of the plurality of sensors for each existing track and the observation identifications of the plurality of sensors for each first target; determining the approximation degree between the candidate track and the candidate target according to the observation identifications of the plurality of sensors for the candidate track and the observation identifications of the plurality of sensors for the candidate target; and carrying out track association on the candidate track and the candidate target according to the approximation degree to obtain a first association result.
Optionally, the matching module is further configured to construct an observation identifier set corresponding to each existing track according to the observation identifiers of the plurality of sensors for each existing track; if a plurality of observation identification sets exist, a target observation identification set comprising observation identifications of a plurality of sensors aiming at a first target is obtained, an existing track corresponding to the target observation identification set is taken as a candidate track, and the first target is obtained as a candidate target.
Optionally, the matching module is further configured to determine a ratio of observation identifiers of the plurality of sensors for the candidate targets in the observation identifier set corresponding to the candidate track, as an identifier ratio corresponding to the candidate target under the candidate track; and determining the approximation degree between the candidate track and the candidate target according to the mark duty ratio corresponding to the candidate target under the candidate track and the distance between the candidate track and the candidate target.
Optionally, the matching module is further configured to divide an identification duty ratio corresponding to the candidate target under the candidate track by a distance between the candidate track and the candidate target to obtain an approximation degree between the candidate track and the candidate target.
Optionally, the track association matching result further comprises a second association result; and the matching module is also used for determining that a first track which is not associated with the first target exists in the plurality of existing tracks and a third target which is not associated with the existing track exists in the plurality of first targets according to the first association result, and carrying out track association on the first track and the third target according to the distance between the first track and the third target to obtain a second association result.
Optionally, the track association matching result further comprises a third association result; the matching module is further configured to determine, if a second track that is not associated with the first target exists in the plurality of existing tracks and a fourth target that is not associated with the existing track exists in the plurality of first targets according to the first association result and the second association result, and associate, for each fourth target, the fourth target to a second track that is closest to the fourth target, to obtain a third association result; and if the fifth target which is not associated with the existing track exists in the plurality of second targets according to the first association result, the second association result and the third association result, ignoring the fifth target.
Optionally, the plurality of first targets are observed at observation points; the matching module is further configured to determine, if the measurement point is observed to be located in the target area of the intersection, according to the first association result and the second association result, that a second track which is not associated with the first target exists in the plurality of existing tracks, and that a fourth target which is not associated with the existing track exists in the plurality of first targets, and for each fourth target, associate the fourth target to a second track closest to the fourth target in distance, so as to obtain a third association result.
In a third aspect, an embodiment of the present application provides an electronic device, including a processor and a memory; one or more programs are stored in the memory and configured to be executed by the processor to implement the methods described above.
In a fourth aspect, embodiments of the present application provide a computer readable storage medium having program code stored therein, wherein the program code, when executed by a processor, performs the method described above.
In a fifth aspect, embodiments of the present application provide a computer program product or computer program comprising computer instructions stored in a computer readable storage medium. The processor of the electronic device reads the computer instructions from the computer readable storage medium and executes the computer instructions to cause the electronic device to perform the method described above.
In the track processing method, the device, the electronic equipment and the storage medium provided by the embodiment of the application, if a plurality of existing tracks have target existing tracks which are not related to a first target, determining that a second target which is related to the target existing tracks before the target time is not observed at a prediction time point, predicting the target existing tracks corresponding to the second target, and determining the prediction confidence of the prediction result according to the prediction course angle corresponding to the second target at the prediction time point, the course angle corresponding to the last position point in the target existing tracks and the number of target prediction position points in the target existing tracks, wherein the prediction confidence of the prediction result can accurately represent the accuracy of the prediction result of the second target at the prediction time point, and when the prediction confidence is larger than or equal to a confidence threshold, the prediction result of the second target at the prediction time point is accurate, the prediction position point of the second target at the prediction time point is more approximate to the actual observation position point of the second target at the prediction time point, and the prediction position point of the second target at the prediction time point can be replaced by the prediction position point of the second target at the prediction time point, so that the prediction confidence of the prediction result is higher than the actual observation position of the second target at the prediction time point of the second target existing tracks.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the description of the embodiments will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present application, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 shows a schematic diagram of an application scenario to which an embodiment of the present application is applicable;
FIG. 2 is a flow chart of a track processing method according to an embodiment of the present application;
FIG. 3 is a schematic diagram of an existing trajectory of a second object at a predicted point in time according to an embodiment of the present application;
FIG. 4 is a flow chart of a track processing method according to yet another embodiment of the present application;
FIG. 5 shows a schematic diagram of a track in the related art;
FIG. 6 is a schematic diagram of a first associated track in an embodiment of the application;
FIG. 7 is a flow chart of a track processing method according to still another embodiment of the present application;
FIG. 8 is a schematic diagram of a trace after omitting a fifth object in an embodiment of the application;
FIG. 9 is a schematic diagram of a target area in an embodiment of the application;
FIG. 10 is a schematic diagram of a trace processing procedure in an embodiment of the application;
FIG. 11 is a block diagram of a track processing device according to an embodiment of the present application;
fig. 12 shows a block diagram of an electronic device for performing a trajectory processing method according to an embodiment of the present application.
Detailed Description
The following description of the embodiments of the present application will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present application, but not all embodiments. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the application, are within the scope of the application in accordance with embodiments of the present application.
In the following description, the terms "first", "second", and the like are merely used to distinguish between similar objects and do not represent a particular ordering of the objects, it being understood that the "first", "second", or the like may be interchanged with one another, if permitted, to enable embodiments of the application described herein to be practiced otherwise than as illustrated or described herein.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of the application only and is not intended to be limiting of the application.
It should be noted that: references herein to "a plurality" means two or more. "and/or" describes an association relationship of an association object, meaning that there may be three relationships, e.g., a and/or B may represent: a exists alone, A and B exist together, and B exists alone. The character "/" generally indicates that the context-dependent object is an "or" relationship.
The application discloses a track processing method, a track processing device, electronic equipment and a storage medium, and relates to the technology of Internet of vehicles.
The intelligent transportation system (Intelligent Traffic System, ITS), also called intelligent transportation system (Intelligent Transportation System), is a comprehensive transportation system which uses advanced scientific technology (information technology, computer technology, data communication technology, sensor technology, electronic control technology, automatic control theory, operation study, artificial intelligence, etc.) effectively and comprehensively for transportation, service control and vehicle manufacturing, and enhances the connection among vehicles, roads and users, thereby forming a comprehensive transportation system for guaranteeing safety, improving efficiency, improving environment and saving energy.
The intelligent vehicle-road cooperative system (Intelligent Vehicle Infrastructure Cooperative Systems, IVICS), which is simply called a vehicle-road cooperative system, is one development direction of an Intelligent Transportation System (ITS). The vehicle-road cooperative system adopts advanced wireless communication, new generation internet and other technologies, carries out vehicle-vehicle and vehicle-road dynamic real-time information interaction in all directions, develops vehicle active safety control and road cooperative management on the basis of full-time idle dynamic traffic information acquisition and fusion, fully realizes effective cooperation of people and vehicles and roads, ensures traffic safety, improves traffic efficiency, and forms a safe, efficient and environment-friendly road traffic system. The method can be applied to an intelligent traffic system or an intelligent vehicle-road cooperative system, and is used for observing the vehicle (target) by combining sensor information determined by road test perception, and carrying out track association aiming at the existing track of the vehicle and the observed position point of the newly observed vehicle so as to update the real-time track of the vehicle.
At present, an observation target observed at an observation time point can be obtained by observing a sensor pair, an existing track of the observation time point is obtained, and the observation target and the existing track are associated to obtain a new existing track. When the existing track is not matched with the observed target, the existing target associated with the existing track is considered to be lost at the observed time point, and the existing track which is not matched with the observed target is not used for carrying out subsequent track association process. The sensor or environmental factors cause that the observation time point cannot observe the position point of the existing target, so that the existing track of the existing target cannot be matched with the observation target at the observation time point, the existing target is not lost, and the follow-up track association process is not performed by utilizing the existing track which is not matched with the observation target, so that the track quality obtained by carrying out track association in the follow-up process is poor.
In the related art, if an existing track at a certain time point is not associated with an observed target, the existing track can be predicted to obtain a predicted position point of the existing target associated with the existing track at the observed time point, and then the association between the existing track and the predicted position point is established to obtain the existing track of the existing target at the time point. However, the accuracy of predicting the target prediction using the existing method is low, resulting in a low accuracy of the existing trajectory constructed from the predicted location points.
In view of this, in the track processing method, the device, the electronic equipment and the storage medium provided by the embodiment of the application, if there are existing tracks of targets which are not associated with the first target in the existing tracks, it is determined that a second target associated with the existing tracks of the targets before the target time is not observed at the prediction time point, the existing tracks of the targets corresponding to the second target are predicted, and according to the predicted course angle corresponding to the second target at the prediction time point, the course angle corresponding to the last position point in the existing tracks of the targets and the number of target predicted position points in the existing tracks of the targets, the prediction confidence of the prediction results is determined, the accuracy of the prediction results of the second target at the prediction time point can be accurately represented, and when the prediction confidence is greater than or equal to the confidence threshold, the prediction results of the second target at the prediction time point are indicated to be accurate, the prediction position point of the second target at the prediction time point is more approximate to the actual observation position point of the second target at the prediction time point, and the prediction position point of the second target at the prediction time point can be replaced by the prediction position point of the second target at the prediction time point, so that the prediction confidence of the prediction results of the second target at the prediction time point is higher than the actual observation position point of the target at the prediction time point.
As shown in fig. 1, an application scenario to which the embodiment of the present application is applicable includes a terminal 20 and a server 10, where the terminal 20 and the server 10 are connected through a wired network or a wireless network. The terminal 20 may be a smart phone, a tablet computer, a notebook computer, a desktop computer, a smart home appliance, a vehicle-mounted terminal, an aircraft, a wearable device terminal, a virtual reality device, and other terminal devices capable of page presentation, or other applications (e.g., instant messaging applications, shopping applications, search applications, game applications, forum applications, map traffic applications, etc.) capable of invoking page presentation applications. In other embodiments, the terminal 20 may also be a device capable of location information acquisition, such as a laser radar, an image acquisition device, etc., which is not specifically limited herein.
The server 10 may be an independent physical server, a server cluster or a distributed system formed by a plurality of physical servers, or a cloud server providing cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDNs (Content Delivery Network, content delivery networks), basic cloud computing services such as big data and artificial intelligent platforms, and the like. The server 10 may be used to provide services for applications running at the terminal 20.
The terminal 20 may send a plurality of existing tracks before the target time point and a plurality of first targets observed at the target time point to the server 10, where the server 10 performs the track processing method according to the present application on the plurality of existing tracks before the target time point and the plurality of first targets observed at the target time point to obtain a new existing track corresponding to the second target at the predicted time point and a track association matching result after performing track association on the predicted position point corresponding to the predicted time point and the target existing track corresponding to the second target at the predicted time point, and the server 10 returns the new existing track corresponding to the second target at the predicted time point and the track association matching result to the terminal 20.
In another embodiment, the terminal 20 may be configured to perform the track processing method of the present application to obtain a new existing track corresponding to the second target at the predicted time point and a track association matching result.
For convenience of description, in the following embodiments, track processing is described as an example performed by an electronic device.
Referring to fig. 2, fig. 2 is a flowchart illustrating a track processing method according to an embodiment of the present application, where the method may be applied to an electronic device, and the electronic device may be the terminal 20 or the server 10 in fig. 2, and the method includes:
S110, performing association matching on a plurality of existing tracks before the target time point and a plurality of first targets observed at the target time point to obtain a track association matching result.
In the present application, a plurality of first targets observed at a target time point may be determined by data detected by a plurality of sensors, which may refer to devices for observing surrounding environmental information, for example, the sensors may be cameras, ultrasonic radars, and lidar, a positioning module, and the like. The first object refers to an object observed at an object time point, which may be a vehicle, or other device capable of being positioned, without being specifically limited herein. It may be appreciated that the plurality of first targets observed at the target time point at least carries position information of the first target determined by observation at the target time point (i.e. the observation position point hereinafter), and further may also carry motion information (such as a motion speed, a motion acceleration, a heading angle, etc. which are not specifically limited herein) of the first target determined by observation at the target time point.
The target time point may be any one of observation time points, and the target may be a vehicle, and any one of the observation time points at which the vehicle travels on the road s1 is taken as one target time point.
The observation time point refers to a time when an observed object (target) observes the target through a plurality of sensors, the observation position point refers to a position point where the target is determined by fusion sensing of sensor information at the observation time point, and in practice, motion information of the target, such as a motion speed, a motion acceleration, a heading angle, and the like, can also be determined by fusion of a plurality of sensor information.
The fusion sensing refers to sensing and identifying targets in a scene by using sensors distributed in multiple directions, the types can include but are not limited to acquisition cameras, millimeter wave radars and the like, and the purposes of complementation of sensing and detecting ranges and complementation of different sensor capacities are achieved by fusing multiple sensors and multiple sensor fields of view.
And at any observation time point, observing the environment around the electronic equipment through a plurality of sensors to obtain sensor information of each sensor aiming at the target, and then fusing the sensor information of each sensor aiming at the target to obtain the observation position point of the target at the observation time point. For example, the sensor includes a camera and a lidar, and for an observation time point t1, an image of a target a1 and an image of a target a2 are observed by the camera, radar information of the target a1 and radar information of a2 are observed by the lidar, the image of a1 and the radar information are fused to obtain an observation position point of a1 at t1, and the image of a2 and the radar information are fused to obtain an observation position point of a2 at t 1.
It should be noted that, in order to save computing resources and energy consumption, the sensor does not need to observe the environment around the electronic device at any moment, and may observe the environment around the electronic device at preset time intervals to obtain an observed location point of the observed object, that is, the observed time point may not be a continuous time point.
At the target time point, sensor information of each first target is acquired through a plurality of sensors, the plurality of sensors fuse the sensor information of each first target to obtain observation information (such as the observation position point in the above, and further including motion information) of each first target observed by the plurality of sensors, so that the plurality of sensors observe the plurality of first targets at the target time point.
The track corresponding to the target is a track obtained by sequentially connecting the position points of the target at different time points according to the time sequence. Each existing track before the target time point is obtained by sequentially connecting observation position points indicated by observation information of the same target before the target time point, and each existing track comprises at least one position point; of course, in other embodiments, if the target is predicted before the target time point, the existing track may further include a predicted position point predicted for the target, where the motion information of the target associated with the predicted position point is motion information (for example, heading angle) determined by prediction.
The initial position points of the multiple existing targets can be obtained at the initial moment of generating the track, and the observed target (or predicted target) and the initial position points of the multiple existing targets are subjected to track association at the next observation time point to obtain the track of the multiple existing targets at the observation time point; and carrying out track association on the observed target or the predicted target and tracks of a plurality of existing targets at each subsequent observation time point to obtain the tracks of the plurality of existing targets at the observation time point.
The predicted time point refers to a time point of a predicted position point of an existing target according to a position point of the existing target, and the position point of the target predicted by filtering refers to the predicted position point of the existing target. The predicted position point of the existing target (or the existing target associated with the existing trajectory) at the predicted time point may be predicted by a kalman filtering method based on the existing position point (or the existing trajectory) of the existing target before the predicted time point. In order to save the computing resources and the energy consumption, the prediction time point is not a continuous time point, and the prediction is performed only when the prediction time point is reached.
The filtering means that track association and filtering are carried out on the continuous target observation information containing noise after fusion, and the real target three-dimensional track is estimated. The filtering mode includes, but is not limited to, kalman filtering, window smoothing filtering, etc.
Track association refers to associating an existing location point (or an existing track) of an existing target with an observed target to connect the existing location point (or the existing track) of the existing target satisfying the association condition with the observed target as a new track. Meeting the association condition may mean that a distance between an existing location point (or an existing trajectory) of an existing target and an observed target is less than a preset distance.
In this embodiment, for any time point, the relative distance between the existing position point (or the last position point in the existing track) before the time point and the position point where the observed target at the time point is located may be determined according to the existing position point (or the last position point in the existing track) of the existing target before the time point and the position point where the observed target at the time point is located, and if the relative distance is smaller than the preset relative distance, the correlation condition is determined to be satisfied. The relative distance herein may refer to euclidean distance, cosine distance, square of euclidean distance, and the like. The preset relative distance may be a value set based on the requirement, which is not limited by the present application. The relative distance of a track from a location point may refer to the relative distance of the last location point in the track from that location point.
For example, the target time point is preceded by 2 time points, the electronic device observes observation position points of 3 targets at the 1 st time point, observes observation position points of 3 targets at the 2 nd time point, and correlates the observation position points of 3 targets observed at the 1 st time point with the observation position points of 3 targets observed at the 2 nd time point to obtain 3 tracks, and the 3 tracks are taken as 3 existing tracks before the target time point.
As described above, the observation position points of the plurality of first targets may be observed by the plurality of sensors, and the plurality of existing trajectories before the target time point and the respective observation position points of the plurality of first targets observed at the target time point may be associated with each other to achieve association matching of the plurality of existing trajectories before the target time point and the plurality of first targets observed at the target time point, so as to obtain the trajectory association matching result.
The track association matching result may include at least one association result corresponding to the existing track after association with the first target. For example, the first targets include 3 existing tracks including 4, the 4 existing tracks and the 3 first targets are subjected to association matching, each first target can find the existing track meeting association conditions, and at this time, 3 association results corresponding to the 3 first targets are obtained.
S120, if it is determined that a target existing track which is not associated with the first target exists in the plurality of existing tracks according to the track association matching result, prediction is performed according to the target existing track, a prediction result corresponding to a second target associated with the target existing track before the target time at a prediction time point is determined, and the prediction result comprises a prediction position point and a prediction course angle corresponding to the second target at the prediction time point.
The existing track which is not associated with the first object in the plurality of existing tracks can be determined according to the track association matching result, the existing track which is not associated with the first object in the plurality of existing tracks is used as the object existing track, the object existing track is a certain existing object corresponding to the existing track, and the existing object is used as the second object which is associated with the object existing track.
And the existing tracks of the targets which are not associated with the first target exist in the existing tracks, which indicates that a second target associated with the existing tracks of the targets is not observed at the target time point, and prediction needs to be carried out according to the existing tracks associated with the second target, so that a prediction result of the second target at the prediction time point is obtained. The prediction result of the second target at the prediction time point may be determined by means of kalman filtering. The predicted time point may be a target time point or a time point after the target time point.
For example, the first targets include 3, the existing tracks include 4, the 4 existing tracks and the 3 first targets are subjected to association matching, each first target can find an existing track meeting association conditions, at this time, 3 association results corresponding to the 3 first targets are obtained, at this time, an existing track s2 which is not associated with any one of the 3 first targets is determined as an existing track of a target, and an existing target a5 which is associated with the existing track s2 of the target before the target time point is determined as a second target. And then predicting a predicted result of a5 at a predicted time point according to the existing track s2 of the target.
In some embodiments, motion estimation may be performed according to the principle of kalman filtering according to the position information corresponding to the last position point in the existing track of the target and the motion information associated with the last position point, so as to determine the prediction result corresponding to the second target at the prediction time point. The prediction result includes position information of a predicted position point of the second target corresponding to the predicted time point, and predicted motion information, such as a predicted heading angle, at the predicted time point.
As an embodiment, S120 may include: if the fields of view of the plurality of sensors at the target time point have intersection, determining that the target existing track which is not associated with the first target exists in the plurality of existing tracks according to the track association matching result, predicting according to the target existing track, and determining a prediction result corresponding to a prediction time point of a second target which is associated with the target existing track before the target time point after the target time point.
In general, in a case where there is intersection between the fields of view of the plurality of sensors, a second object associated with an existing track of the object before the point in time of the object appears to flash more frequently, for example, not observed at the previous point in time of observation, but observed at the next point in time, in which case, if no processing is performed, the observed information of the object is associated with the corresponding track in a case where the object is observed, and in a case where the object is not observed, such that disconnection of the tracks corresponding to the same object easily occurs, a situation where the tracks temporarily disappear and reappear occurs, and in a case where the flashing object frequently flashes on the track (reappears after disappearance) if no processing is performed, and the reappear after appearance is serious, the display effect is affected. At this time, a step of determining that there is a target existing track which is not associated with the first target among the plurality of existing tracks according to the track association matching result, predicting according to the target existing track, and determining a prediction result corresponding to a prediction time point of a second target associated with the target existing track before the target time point after the target time point is performed.
In some embodiments, in the case that there is no intersection between the fields of view of the plurality of sensors, the difficulty of observing the second object associated with the existing track of the object is small, and at this time, if the existing track of the object is not associated with any first object, it means that the first object does not include the second object, the second object is lost, i.e. the existing track of the object is discarded, the step S120 is not performed, and a prompt message is output to prompt that the second object disappears and that the existing track of the object associated with the second object is abnormal. Of course, in other embodiments, in the non-flash area, according to the method of the present application, for a target existing track that is not associated with the first target, the target existing track is predicted, a prediction result corresponding to a second target associated with the target existing track before the target time at the prediction time point is determined, and a subsequent process is performed, so as to reduce the occurrence of the flash.
S130, determining the prediction confidence of a prediction result according to a predicted course angle corresponding to a second target at a prediction time point, a course angle corresponding to the last position point in the existing target track and the number of target predicted position points in the existing target track; the target predicted position point refers to the predicted position point located after the last observed position point in the existing track of the target.
The prediction confidence of the prediction result is used for indicating the reliability degree of the prediction result, and the higher the prediction confidence of the prediction result is, the more reliable and the higher the accuracy of the prediction result is, the lower the prediction confidence of the prediction result is, the less reliable and the lower the accuracy of the prediction result is.
The observation position point is a position point determined by targeting sensor information of a plurality of sensors, and the predicted position point is a position point estimated for the target by prediction (for example, prediction by kalman filtering). The predicted position point after the last observed position point in the existing track of the target is taken as the predicted position point of the target. For example, the existing track of the target includes 7 position points, the first four position points are observation position points, the last 3 position points are all prediction position points, and the last 3 position points are all target prediction position points. The number of the target predicted position points in the target existing track is the position points which are continuously predicted and determined after the last observation position point in the target existing track.
In practice, there may be an error between the predicted position point predicted for the target and the actual observed position point of the target, so if the number of target predicted position points in the existing track of the target is large, there may be a large error of the predicted result determined at the target time point for the second target at the predicted time point due to error accumulation, that is, the prediction confidence of the predicted result is low. In other words, the number of target predicted position points in the existing target track and the prediction confidence of the prediction result are in a negative correlation.
It will be appreciated that the direction of movement of the same target at two adjacent time points (with a short time interval) is unlikely to be abrupt. According to the predicted course angle of the second target corresponding to the predicted time point and the course angle corresponding to the last position point in the existing track of the target, the difference between the predicted course angle and the course angle corresponding to the last position point in the existing track of the target can be determined, if the difference is large, the situation that the movement direction of the second target changes greatly is indicated, and the occurrence probability of the situation is low, so that the prediction confidence of the prediction result is low, namely the difference between the predicted course angle and the course angle corresponding to the last position point in the existing track of the target is in a negative correlation relationship with the prediction confidence of the prediction result.
In some embodiments, the calculation may be performed on the predicted heading angle, the heading angle corresponding to the last position point in the existing track of the target, and the number of target predicted position points in the existing track of the target, to obtain a prediction confidence of the prediction result, where the calculation process may include: calculating the difference between the predicted course angle corresponding to the predicted time point of the second target and the course angle corresponding to the last position point in the existing track of the target; and determining the prediction confidence of the prediction result according to the number and the difference of the target prediction position points. The difference between the predicted heading angle corresponding to the predicted time point and the heading angle corresponding to the last position point in the existing track of the target may refer to an absolute value of a difference between the predicted heading angle corresponding to the predicted time point and the heading angle corresponding to the last position point in the existing track of the target, or may refer to a result after normalization of an absolute value of a difference between the predicted heading angle corresponding to the predicted time point and the heading angle corresponding to the last position point in the existing track of the target.
When the prediction confidence of the prediction result is calculated, the prediction confidence of the prediction result can be determined according to the negative correlation between the prediction confidence of the prediction result and the number of target prediction position points and the negative correlation between the prediction confidence of the prediction result and the difference. In some embodiments, the number of target predicted location points and the absolute value of the difference may be weighted, and the inverse of the weighted result is taken as the prediction confidence of the prediction result; the weighting coefficients corresponding to the number of the target predicted position points and the weighting coefficients corresponding to the absolute values of the differences can be determined through experiments. As an implementation manner, the absolute value of the difference between the predicted course angle corresponding to the predicted time point and the course angle corresponding to the last position point in the existing track of the target can be normalized, and the prediction confidence of the prediction result can be determined according to the normalized value and the number of the predicted position points of the target. In this case, the prediction confidence and the normalized value have a negative correlation, and the prediction confidence and the number of target predicted position points in the target existing trajectory have a negative correlation.
And S140, if the prediction confidence coefficient is greater than or equal to the confidence coefficient threshold value, carrying out track association on the predicted position point corresponding to the predicted time point of the second target and the existing track of the target.
If the prediction confidence is greater than or equal to the confidence threshold, the accuracy of the prediction result corresponding to the second target at the prediction time point is higher, the prediction position point corresponding to the second target at the prediction time point is very close to the actual position point of the second target at the prediction time point, the prediction position point corresponding to the second target at the prediction time point can be used as the actual position point of the second target at the prediction time point, and the prediction position point corresponding to the second target at the prediction time point and the existing track of the target are subjected to track association, so that the obtained existing track of the second target corresponding to the prediction time point.
The confidence threshold may be a value determined based on the specific operation process of the demand and the predicted confidence, and may be set according to the actual demand, which is not particularly limited herein.
If the prediction confidence is smaller than the confidence threshold, the accuracy of the prediction result corresponding to the second target at the prediction time point is lower, the prediction position point corresponding to the second target at the prediction time point is not close to the actual position point of the second target at the prediction time point, the prediction position point corresponding to the second target at the prediction time point cannot be used as the actual position point of the second target at the prediction time point, at the moment, the loss of the second target can be determined, the existing track of the target associated with the second target can be abandoned, and prompt information is output to prompt the disappearance of the second target and the occurrence of abnormality of the existing track of the target associated with the second target.
And after track association is performed on the predicted position point corresponding to the predicted time point and the existing track of the target, the existing track of the second target corresponding to the predicted time point is obtained, the next observation time point after the target time point can be obtained as a new target time point, and the existing track of the second target under the predicted time point and the association result in the track association matching result are taken as the existing track before the new target time point, and the execution returns to S110.
After the step S110 is performed again, the second target may be observed again at a certain observation time point, and when the second target is not associated with the first target, the present application predicts the second target to obtain the present track of the second target at each prediction time point, and when the second target is observed again, the present track of the second target at the last prediction time point before the second target is observed again may be associated with the observation position point of the second target at the observation time point when the second target is observed again, so as to obtain the present track of the second target at the observation time point when the second target is observed again, and the present condition of the present track of the second target corresponding to the second target may be avoided.
Wherein re-observing the second target may refer to: and at any observation time point, the observed position point can be in track association with the existing track corresponding to the second target at the last prediction time point before the observation time point, and the observed position point is determined to be the observed position point of the second target when the second target is observed again.
In addition, when the existing track of the target of the second target is not associated with any one of the first targets, the number of times of predicting the second target may be multiple, the second target may be predicted at a target time point, and in case that the prediction confidence of the prediction result corresponding to the target time point is greater than or equal to the confidence threshold, the second target may be predicted at a next observation time point after the target time point, so that, for any one observation time point, the second target may be predicted at a next observation time point, in other words, each observation time point may be regarded as a prediction time point of an unobserved target, in case that the prediction confidence of the prediction result corresponding to the current observation time point of the second target is greater than or equal to the confidence threshold.
In addition, in the case where there is an intersection between the fields of view of the plurality of sensors, in order to be able to re-observe the second object, the prediction period may be increased, that is, in the process of predicting the second object, the number of prediction time points may be increased (for example, when there is no intersection between the fields of view of the plurality of sensors, the number of prediction time points exceeds 4, prediction is no longer performed, when there is an intersection between the fields of view of the plurality of sensors, the number of prediction time points exceeds 6, prediction is no longer performed), and at the same time, the relative distance threshold corresponding to the relative distances between the existing trajectory and the predicted position point is appropriately increased when the existing trajectory and the predicted position point for each prediction time point are associated (for example, when there is no intersection between the fields of view of the plurality of sensors, the relative distance threshold is 10, and when there is an intersection between the fields of view of the plurality of sensors, the relative distance threshold is 15), so that the success rate at which the predicted position point can be associated with the existing trajectory at the predicted position point is increased.
Under the condition that intersection exists in the fields of view of a plurality of sensors, although the success rate of association can be improved by increasing the prediction time length and the relative distance threshold value, the prediction error can be increased, so that the prediction confidence is introduced, the reliability of the prediction result after the prediction time length and the relative distance threshold value are increased can be accurately indicated through the prediction confidence, when the prediction confidence of the prediction result is larger than or equal to the confidence threshold value, the prediction position point in the prediction result can accurately indicate the actual position point, and the accuracy of the existing track of the second target under the prediction time point is improved.
For example, as shown in a of fig. 3, in the related art, after the existing track 31 (the existing track formed by connecting 1, 2 and 3) of the target is not associated with the first target, it is determined that the second target associated with the existing track 31 of the target is lost, however, when the second target is re-observed, the position point 4 when the second target is re-observed is too far away from the existing track 31 of the target, resulting in that the re-observed second target cannot be associated with the existing track 4 of the target, and thus the track 32 constructed from the observed position point 4 of the second target when the second target is re-observed and the observed position point 5 of the next observation time point is not a continuous track with the existing track 31 of the target, a track break occurs, resulting in poor track quality.
As shown in b in fig. 3, according to the method of the present application, after the existing track 31 (the existing track formed by connecting 1, 2 and 3) of the target is not related to the first target, the second target related to the existing track 31 of the target is predicted to obtain the predicted position point 6 at the predicted time point, the predicted position point 6 is related to the existing track 31 of the target to obtain the new existing track 33 (the existing track formed by connecting 1, 2, 3 and 6), the predicted position point 7 of the next predicted time point is predicted according to the new existing track 33, the predicted position point 7 is related to the existing track 33 to obtain the new existing track 34 (the existing track formed by connecting 1, 2, 3, 6 and 7), when the second target is re-observed, the distance between the position point 4 when the second target is re-observed and the existing track 34 is relatively close, so that the track 32 constructed according to the observed position point 4 and 5 of the second target when the second target is re-observed forms a continuous track, namely, the predicted track at the time point when the second target is re-observed, the existing track is not observed, and the quality of the new track is lost or the existing track is lost temporarily and the new track is not observed.
In this embodiment, if there are existing tracks of targets that are not associated with the first target in the existing tracks, it is determined that a second target associated with the existing track of targets before the target time is not observed at a prediction time point, the existing track of targets corresponding to the second target is predicted, and according to a predicted heading angle corresponding to the second target at the prediction time point, a heading angle corresponding to a last location point in the existing track of targets, and the number of target predicted location points in the existing track of targets, a prediction confidence of the prediction result is determined, and the accuracy of the prediction result of the second target at the prediction time point can be accurately represented, when the prediction confidence is greater than or equal to a confidence threshold, it is indicated that the prediction result of the second target at the prediction time point is accurate, the prediction location point of the second target at the prediction time point is closer to an actual observation location point of the second target at the prediction time point, and the prediction location point of the second target at the prediction time point can replace the actual observation location point of the second target at the prediction time point, so that the prediction location point of the second target at the prediction time point and the existing track of targets corresponding to the second target are associated with each other. According to the method provided by the application, the situation that the track temporarily disappears and reappears in the related technology can be effectively solved, the flash can be effectively restrained, and the probability of track flash in the track fusion process is reduced.
Meanwhile, in the application, under the condition that intersection exists among the visual fields of a plurality of sensors, the prediction duration and the relative distance threshold value in the track association process are increased, so that the success rate of track association of the existing track of the target and the predicted position point of the second target at the prediction time point is improved. In addition, by the method, the condition that the track of the second target is discontinuous due to track flash of the second target is avoided, and the track quality is improved.
Referring to fig. 4, fig. 4 is a flowchart illustrating a track processing method according to another embodiment of the present application, where the method may be applied to an electronic device, and the electronic device may be the terminal 20 or the server 10 in fig. 2, and in this embodiment, the track association matching result includes a first association result, and the method includes:
s210, determining candidate tracks from the existing tracks and candidate targets from the first targets according to the observation identifications of the sensors for each existing track and the observation identifications of the sensors for each first target.
In this embodiment, the first target is observed by a plurality of sensors, that is, respective sensor information is observed by a plurality of sensors at a target time point, and respective sensor information of the respective sensors is fused to obtain respective observation position points of the plurality of first targets, so that it is determined that the plurality of first targets are observed at the target time.
For each sensor, when the sensor observes the first target, the sensor assigns a respective identifier to the observed first target, and the identifier assigned by the sensor to the observed first target serves as the observed identifier of the sensor for the first target. The observation identifications of the sensors for the first targets may be stored in sensor information of the sensors, and the respective observation identifications of the sensors for the respective first targets may be obtained from the sensor information of the sensors.
For example, the plurality of sensors includes a sensor c1 and a sensor c2, the first target includes a target d1 and a target d2, the observation of the target d1 in the sensor information observed by the sensor c1 is denoted by d11, the observation of the target d1 in the sensor information observed by the sensor c2 is denoted by d12 and the observation of the target d2 is denoted by d21, and the observation of the target d1 in the sensor information observed by the sensor c2 is denoted by d12 and the observation of the target d2 is denoted by d22.
The existing track comprises at least one observation position point corresponding to the existing target, and one observation position point in the existing track is used as one existing observation position point. For each existing observation position point corresponding to the existing target, obtaining the observation identification of each sensor for the existing target (when each sensor observes the existing target, the sensor information comprises the existing target allocation observation identification, and the existing target allocation observation identification can be directly obtained and stored), and summarizing the observation identifications of each sensor for all the existing observation position points of the existing target to obtain the observation identification of each sensor for the existing target. The observation identifications of the plurality of sensors aiming at the existing target are summarized to be used as the observation identifications of the plurality of sensors aiming at the existing track corresponding to the existing target.
After the observation identifications of the plurality of sensors for each existing track and the observation identifications of the plurality of sensors for each first target are obtained, candidate tracks may be determined from the plurality of existing tracks and candidate targets may be determined from the plurality of first targets according to an association relationship between the observation identifications of the plurality of sensors for each existing track and the observation identifications of the plurality of sensors for each first target (may refer to an inclusion relationship between the observation identifications of the plurality of sensors for each existing track and the observation identifications of the plurality of sensors for each first target).
As an embodiment, S210 may include: constructing an observation identification set corresponding to each existing track according to the observation identifications of the plurality of sensors aiming at each existing track; if a plurality of observation identification sets exist, a target observation identification set comprising observation identifications of a plurality of sensors aiming at a first target is obtained, an existing track corresponding to the target observation identification set is taken as a candidate track, and the first target is obtained as a candidate target.
The observation identifications of the plurality of sensors for each existing track can be summarized to obtain an observation identification set corresponding to each existing track. For example, the sensor includes a sensor c1 and a sensor c2, the existing trajectory g1 includes an existing observation position point w1 and an existing observation position point w2, the sensor c1 has an observation identification bs11 for an existing target at the existing observation position point w1, the sensor c2 has an observation identification bs12 for an existing target at the existing observation position point w1, the sensor c1 has an observation identification bs21 for an existing target at the existing observation position point w2, the sensor c2 has an observation identification bs22 for an existing target at the existing observation position point w2, bs12, bs11, bs21, and bs22 are acquired as a plurality of observation identifications corresponding to the existing trajectory g1, and bs12, bs11, bs21, and bs22 are summarized as a set as an observation identification set corresponding to the existing trajectory g 1.
Traversing each existing track according to the process to obtain an observation identification set corresponding to each existing track, so as to obtain a plurality of observation identification sets, and if a target observation identification set comprising observation identifications of a plurality of sensors aiming at a first target exists in the plurality of observation identification sets, acquiring the existing track corresponding to the target observation identification set as a candidate track, and acquiring the first target as a candidate target.
For example, the set of observation identifications corresponding to the existing track g1 includes bs12, bs11, bs21, and bs22, the set of observation identifications corresponding to the existing track g2 includes bs32, bs31, bs41, and bs42, the set of observation identifications corresponding to the existing track g3 includes bs52, bs51, bs61, and bs62, the set of observation identifications corresponding to the plurality of sensors for the first target bs11 and bs22 determines that the set of observation identifications corresponding to the existing track g1 includes the plurality of observation identifications corresponding to the first target sensors, and at this time, determines the first target as a candidate target, and determines the existing track g1 as a candidate track.
All the first objects and the existing trajectories are traversed according to the above procedure, all candidate objects and candidate trajectories are determined, and then the following step S220 is continued.
S220, determining the approximation degree between the candidate track and the candidate target according to the observation identifications of the plurality of sensors for the candidate track and the observation identifications of the plurality of sensors for the candidate target.
The proximity between each candidate track and each candidate target may be determined from an association between the observation identifications of the plurality of sensors for each candidate track and the observation identifications of the plurality of sensors for each candidate target. The association relationship between the observation identifications of the plurality of sensors for each candidate trajectory and the observation identifications of the plurality of sensors for each candidate target may refer to an inclusion relationship between the observation identifications of the plurality of sensors for each candidate trajectory and the observation identifications of the plurality of sensors for each candidate target.
For example, the candidate targets include hm1 and hm2, the candidate trajectories include hg1 and hg2, and the proximity between hm1 and hg1 is determined according to the association relationship between the observation identifications of the plurality of sensors for hm1 and the observation identifications of the plurality of sensors for hg 1. The method comprises the steps of determining the approximation degree between hm2 and hg1 according to the incidence relation between the observation identifications of the sensors for hm2 and the observation identifications of the sensors for hg1, determining the approximation degree between hm1 and hg2 according to the incidence relation between the observation identifications of the sensors for hm1 and the observation identifications of the sensors for hg2, and determining the approximation degree between hm2 and hg2 according to the incidence relation between the observation identifications of the sensors for hm2 and the observation identifications of the sensors for hg 2.
Optionally, S220 may include: determining the duty ratio of observation identifications of a plurality of sensors aiming at candidate targets in an observation identification set corresponding to the candidate track, and taking the duty ratio as the identification duty ratio corresponding to the candidate targets under the candidate track; and determining the approximation degree between the candidate track and the candidate target according to the mark duty ratio corresponding to the candidate target under the candidate track and the distance between the candidate track and the candidate target.
And determining the duty ratio of the observation identifications of the plurality of sensors for the candidate targets in the observation identification set corresponding to each candidate track as the identification duty ratio corresponding to the candidate targets under each candidate track. For example, the candidate target includes hm1, the candidate track includes hg1 and hg2, the ratio zb1 of the observed identity of the candidate target hm1 in the observed identity set of hg1 is determined as the identity ratio of hm1 under hg1, and the ratio zb2 of the observed identity of the candidate target hm1 in the observed identity set of hg2 is determined as the identity ratio of hm1 under hg 2.
The identification duty ratio of each candidate target under each candidate track is obtained, and the approximation degree between each candidate track and each candidate target can be determined according to the identification duty ratio corresponding to the candidate target under the candidate track and the distance between the candidate track and the candidate target. For example, the candidate target includes hm1, the candidate track includes hg1 and hg2, the mark ratio of hm1 under hg1 is zb1, the approximation degree between hg1 and hm1 is determined according to zb1 and the Euclidean distance between hg1 and hm1, the mark ratio of hm1 under hg2 is zb2, and the approximation degree between hg2 and hm1 is determined according to zb2 and the Euclidean distance between hg2 and hm 1.
As one embodiment, determining the approximation degree between the candidate track and the candidate target according to the mark duty ratio corresponding to the candidate target under the candidate track and the distance between the candidate track and the candidate target includes: and dividing the mark duty ratio corresponding to the candidate target under the candidate track by the distance between the candidate track and the candidate target to obtain the approximation degree between the candidate track and the candidate target. The distance between the candidate track and the candidate target may refer to the euclidean distance between the candidate track and the candidate target, the normalized euclidean distance, or the like.
In this embodiment, the higher the mark duty ratio corresponding to the candidate target under the candidate track, the higher the mark duty ratio of the observed mark of the candidate target observed at the target time point in the observed mark of the candidate track, and the greater the possibility that the candidate target is the target associated with the candidate track, therefore, the mark duty ratio corresponding to the candidate target under the candidate track and the approximation degree between the candidate track and the candidate target are in positive correlation. Meanwhile, the larger the distance between the candidate trajectory and the candidate target is, the less likely the candidate target is the target with which the candidate trajectory is associated, and therefore, the distance between the candidate trajectory and the candidate target, and the approximation degree between the candidate trajectory and the candidate target are in a negative correlation. Therefore, the mark duty ratio corresponding to the candidate target under the candidate track can be divided with the distance between the candidate track and the candidate target to obtain the approximation degree between the candidate track and the candidate target, so that the approximation degree and the mark duty ratio corresponding to the candidate target under the candidate track are in positive correlation, and the distance between the candidate track and the candidate target is in negative correlation with the approximation degree.
And S230, carrying out track association on the candidate track and the candidate target according to the approximation degree to obtain a first association result.
And carrying out track association on each candidate target and each candidate track according to the similarity between each candidate track and each candidate target and a greedy algorithm, at least one first association track, and summarizing all the first association tracks to obtain a first association result. The first association result may be regarded as a track association matching result.
Greedy algorithms are a simpler, faster design technique for some of the best solution problems. The greedy algorithm is characterized in that the greedy algorithm is performed step by step, and is usually used for carrying out optimal selection according to a certain optimization measure based on the current situation, various possible overall situations are not considered, and a great amount of time which is needed to be consumed for finding the optimal solution and exhausting all possible situations is saved. The greedy algorithm adopts a top-down method to make successive greedy selections in an iterative manner, and each time greedy selection is made, the required problem is reduced to a sub-problem with smaller scale, and an optimal solution of the problem can be obtained through greedy selection of each step. Although it is guaranteed that a locally optimal solution is obtained at each step, the global solution thus generated is sometimes not necessarily optimal, so that the greedy algorithm does not need to backtrack.
For example, the candidate track with the largest similarity may be track-associated with the candidate target to obtain a first associated track, then, from the remaining candidate tracks and candidate targets, the candidate track with the largest similarity is selected again to be track-associated with the candidate target to obtain a further first associated track, and the process is performed until all the candidate targets or all the candidate tracks are traversed (all the candidate targets are traversed, all the candidate tracks remain, or all the candidate tracks are traversed, and all the candidate targets remain), so as to obtain at least one first associated track, and all the obtained first associated tracks are summarized as the first associated result.
For example, as shown in fig. 5, in the related art, the first correlation result is determined by not using the observation identifications of the plurality of sensors for each existing track and the observation identifications of the plurality of sensors for each first target, but directly correlating the first correlation result with the relative distance between the existing track and the first target, the target positions outputted in the fusion are transformed to a unified ground plane coordinate system, and errors may be introduced in the coordinate transformation process, so that the track 51 corresponding to the target 1 and the track 52 corresponding to the target 2 are not continuous any more, and a track ID fracture (the situation that the same track is broken) is generated. However, in reality, the target 1 and the target 2 are the same target, and the track 51 and the track 52 belonging to the same target should be continuous, but discontinuous in fig. 5, that is, the accuracy of the correlation directly through the relative distance between the existing track and the first target is low in the related art.
As shown in fig. 6, by the method of the present application, the observation marks of the plurality of sensors for each existing track and the observation marks of the plurality of sensors for each first target determine that the target 3 is the same target as the target 4, so as to determine that the track 61 of the target 3 has a larger approximation degree with the first position point 601 in the track 62 of the target 4, the track 61 of the target 3 and the first position point 601 in the track 62 of the target 2 can be associated with each other, the track 61 and the track 62 are combined into one continuous track, namely, a first associated track, so that the accuracy of the second associated track determined by the plurality of sensors for each existing track and the observation mark of the plurality of sensors for each first target is higher, and the condition that the track ID of the first associated track breaks is greatly reduced, so that the accuracy of the first track association result determined according to the first associated track is higher.
S240, if it is determined that the target existing track which is not associated with the first target exists in the plurality of existing tracks according to the track association matching result, prediction is performed according to the target existing track, a prediction result corresponding to a second target associated with the target existing track before the target time at a prediction time point is determined, and the prediction result comprises a prediction position point and a prediction course angle corresponding to the second target at the prediction time point.
S250, determining the prediction confidence of the prediction result according to the predicted course angle of the second target corresponding to the prediction time point, the course angle corresponding to the last position point in the existing track of the target and the number of the target predicted position points in the existing track of the target.
And S260, if the prediction confidence coefficient is greater than or equal to the confidence coefficient threshold value, carrying out track association on the predicted position point corresponding to the predicted time point of the second target and the existing track of the target.
The descriptions of S240-S260 refer to the descriptions of S120-S140 above, and are not repeated here.
In this embodiment, according to the observation identifications of the plurality of sensors for each existing track and the observation identifications of the plurality of sensors for each first target, the candidate target and the candidate track are determined, the approximation degree between the candidate track and the candidate target is determined, and the approximation degree between the candidate target and the candidate track can accurately represent the association degree between the candidate track and the candidate target, so that the accuracy of the first association result obtained according to the approximation degree is higher, the accuracy of the target candidate track and the second target determined according to the track association matching result including the first association result is higher, and the accuracy of the existing track obtained after the track association is performed on the predicted position point corresponding to the predicted time point and the target existing track of the second target can be further improved.
In this embodiment, through transforming the position information observed by each of the plurality of sensors to the observed position point of the target obtained after unifying the ground plane coordinate system, however, errors may be introduced in the coordinate transformation process, so that the calculation of the relative distance between the determined observed position point and the track exceeds the relative distance threshold value and the association fails, so that the track of the same target is broken.
Referring to fig. 7, fig. 7 shows a flowchart of a track processing method according to still another embodiment of the present application, where the method may be applied to an electronic device, and the electronic device may be the terminal 20 or the server 10 in fig. 2, and in this embodiment, the track association matching result further includes a second association result and a third association result, and the method includes:
S310, acquiring a first association result.
The description of S310 refers to the descriptions of S210 to S230 above, and will not be repeated here.
And S320, if the first track which is not associated with the first target exists in the plurality of existing tracks and the third target which is not associated with the existing track exists in the plurality of first targets according to the first association result, carrying out track association on the first track and the third target according to the distance between the first track and the third target, and obtaining a second association result.
For example, the first object includes m1, m2, m3, m4, and m5, the existing trajectories include yg1, yg2, yg3, yg4, and yg5, the candidate objects are determined to be m1 and m2, the candidate trajectories include yg1, yg2, and yg3, the trajectories of m1 and yg1 are associated by the method of S210-S230, the trajectories of m2 and yg2 are associated, at this time, m3, m4, and m5 are determined to be the third object, and yg3, yg4, and yg5 are determined to be the first trajectory.
And determining the distance between each first track and each third target, carrying out track association on each first track and each third target according to a greedy algorithm to obtain at least one second association track, and summarizing the obtained second association tracks as second association results. The distance between the first track and the third target may be a euclidean distance, a cosine distance, a normalized euclidean distance, or the like.
S330, if a second track which is not associated with the first target exists in the plurality of existing tracks and a fourth target which is not associated with the existing track exists in the plurality of first targets according to the first association result and the second association result, associating the fourth target with the second track which is closest to the fourth target for each fourth target, and obtaining a third association result; and if the fifth target which is not associated with the existing track exists in the plurality of second targets according to the first association result, the second association result and the third association result, ignoring the fifth target.
As described above, according to the first association result, determining a first track which is not associated with the first object in the plurality of existing tracks, and determining a third object which is not associated with the existing track in the plurality of first objects; and determining a fourth target which is not associated with the third target in the third targets according to the second association result, and determining a second track which is not associated with the third target in the first track.
And aiming at each fourth target, associating the fourth target with a second track closest to the fourth target to obtain a third association track, traversing all fourth targets or all second tracks (when the fourth target completes traversing and the second tracks remain, the second tracks are not traversed any more, when the second tracks complete traversing and the fourth targets remain, the second tracks are smaller than the fourth targets, the third association tracks are not traversed any more, and a third association result is obtained by summarizing all third association tracks. The distance between the second track and the fourth target may refer to a euclidean distance, a cosine distance, or a square of the euclidean distance.
The first association result, the second association result, and the third association result may be summarized as track association matching results.
As described above, according to the first association result, determining a first track which is not associated with the first object in the plurality of existing tracks, and determining a third object which is not associated with the existing track in the plurality of first objects; and then according to the second association result, determining a fourth target which is not associated with the third target in the third targets, and determining a second track which is not associated with the third target in the first track, and then according to the third association result, determining a fourth target which is not associated with the second track in the fourth targets, wherein the number of the fourth targets is larger than that of the second tracks, and taking the fourth target which is not associated with the second track in the fourth targets as a fifth target, and neglecting the fifth target.
And carrying out track association on a first target and an existing track, after the first association result, the second association result and the third association result are obtained, traversing all first targets and all existing tracks, and when redundant first targets exist after the first targets are associated with all existing tracks, the redundant first targets are the fifth targets, and the fifth targets are ignored. Similarly, when there is an unnecessary existing track after the first object is associated with all the existing tracks, the unnecessary existing track is the existing track of the object, and the subsequent step S340 is performed according to the existing track of the object.
Since the object does not suddenly appear and disappear, when the first object is observed, the first object is observed to be an existing object which has occurred once, and if the object which is not associated with the existing track is observed to be a split ghost of the existing object (ghost caused by different observations of the object due to the association failure of the method), that is, if a fifth object which is not associated with the existing track exists in the first object, the fifth object is the split ghost of the existing object.
The redundant fifth target may be the split ghost of the first target, and the fifth target is ignored, so that the effect of splitting suppression can be realized, and the occurrence of unclear track caused by the split ghost is avoided. In the related art, the redundant fifth object is not ignored, and the resulting track is shown as a in fig. 8, and the ghost 801 (redundant fifth object) is split in the track 81, resulting in a lower definition of the track 81. After determining the redundant fifth object according to the present application, ignoring the fifth object, the resulting track is shown as b in fig. 8, where the track 82 has no split ghost, the definition of the track 82 is higher, and the track is more accurate.
As one embodiment, the plurality of first targets are observed at an observation point; if, according to the first association result and the second association result, a second track which is not associated with the first target exists in the plurality of existing tracks and a fourth target which is not associated with the existing track exists in the plurality of first targets, for each fourth target, associating the fourth target to a second track which is closest to the fourth target, and obtaining a third association result, including: if the measuring point observation is located in the target area of the intersection, and according to the first association result and the second association result, determining that a second track which is not associated with the first target exists in the plurality of existing tracks and a fourth target which is not associated with the existing track exists in the plurality of first targets, and for each fourth target, associating the fourth target to a second track which is closest to the fourth target in distance to obtain a third association result.
The target region may refer to a region within the intersection, as shown in fig. 9, the target region of the intersection is an ROI (region of interest ) framed by a rectangular frame 901 in the center of the intersection.
An intersection may refer to an intersection: refers to a vehicle passing intersection in an urban scene, and a typical intersection is a standard intersection. But the method can also be used for T-shaped intersections, L-shaped intersections and the like, and is not limited to specific intersection types.
In general, the targets in the intersection target area are not suddenly appeared and disappeared, so when the observation point of the first target is observed to be in the target area, the first target is observed to be the existing target which has been appeared from the observation point, if the target which is not associated with the existing track is observed to be the split ghost of the existing target (the failure of association of the previous method causes the ghost caused by different observation of the target), that is, if there is the fifth target which is not associated with the existing track in the first target, the fifth target is the split ghost of the existing target, and therefore, when the observation point is located in the intersection target area, the process of S330 can be performed to determine the fifth target and ignore the fifth target to avoid that the fifth target which is the split ghost affects the definition and accuracy of the track.
S340, if it is determined that the target existing track which is not associated with the first target exists in the plurality of existing tracks according to the track association matching result, prediction is performed according to the target existing track, a prediction result corresponding to a second target associated with the target existing track before the target time at a prediction time point is determined, and the prediction result comprises a prediction position point and a prediction course angle corresponding to the second target at the prediction time point.
S350, determining the prediction confidence of the prediction result according to the predicted course angle of the second target corresponding to the prediction time point, the course angle corresponding to the last position point in the existing track of the target and the number of the target predicted position points in the existing track of the target.
S360, if the prediction confidence coefficient is greater than or equal to the confidence coefficient threshold value, track association is carried out on the predicted position point corresponding to the predicted time point of the second target and the existing track of the target.
The descriptions of S340 to S360 refer to the descriptions of S120 to S140 above, and are not repeated here.
In this embodiment, the second association result and the third association result are obtained by performing two-time track association on the third target and the first track, and track association is completed on all the third targets or all the first tracks through two-time track association, and the fifth target determined according to the first association result, the second association result and the third association result is ignored, so that the influence of the fifth target determined as a split ghost on the track is avoided, and the definition and accuracy of the track are improved.
In order to more clearly explain the technical solution of the present application, the track processing method of the present application is explained below in conjunction with an exemplary scenario, in which the electronic device is a vehicle, and the vehicle includes 3 sensors cg1, cg2 and cg3, and the existing track at the target time point includes an existing track yyg1 corresponding to the existing target yy1, an existing track yyg2 corresponding to the existing target yy2, an existing track yyg3 corresponding to the existing target yy3, an existing track yyg4 corresponding to the existing target yy4, an existing track yyg5 corresponding to the existing target yy5, and an existing track yyg6 corresponding to the existing target yy 6.
As shown in fig. 10, according to the observation of the surrounding environment of the vehicle by the three sensors at the target time point, sensor information of each of cg1, cg2 and cg3 is obtained, and fusion matching is performed on the sensor information of cg1, cg2 and cg3, so that first targets including dgc1, dgc2, dgc3, dgc4 and dgc are obtained.
Then track re-association is performed: according to the observation marks of cg1, cg2 and cg3 for dgc1, dgc2, dgc3, dgc4 and dgc5 and the observation marks of cg1, cg2 and cg3 for yyg1, yyg, yyg3, yyg4, yyg5 and yyg6, candidate targets are determined to be cg1 and cg2, candidate trajectories are determined to be yyg1 and yyg, then according to the observation marks of cg1, cg2 and cg3 for cg1 and cg2 and the observation marks of cg1, cg2 and cg3 for yyg and yyg2, the approximations of cg1 and cg2 for yyg1 and yyg2 are determined, and according to the approximations, cg1 and yyg are associated to obtain a first association trajectory, and cg2 and yyg2 are associated to obtain a first association trajectory.
Then, target track association is performed: according to the distances dgc, dgc4 and dgc5 from yyg3, yyg4, yyg5 and yyg6, respectively, dgc3 and yyg3 are associated, resulting in a second associated track.
Then, cleavage inhibition was performed: dgc4 is associated with track yyg4 closest to it of yyg, yyg5 and yyg6 to obtain a third associated track, dgc5 is associated with track yyg closest to it of yyg5 and yyg6 to obtain a third associated track, at which point there is no redundant first object and only one existing track yyg6 remains.
Finally, flash suppression is performed: according to the predicted result corresponding to the existing target yy6 at the predicted time point after the predicted target time point of the existing track yyg, determining that the predicted confidence of the predicted result corresponding to the existing target yy6 is higher than a confidence threshold according to the predicted result corresponding to the existing target yy6, and performing track association on the predicted position point in the predicted result corresponding to the existing target yy6 and the existing track yyg to obtain the existing track of the existing target yy6 at the predicted time point.
Output track: the two first associated tracks, the one second associated track, the two third associated tracks and the one existing track of the existing target yy6 under the predicted time point can be summarized to obtain all associated tracks, and all associated tracks are output.
In the scene, track re-association, target track association, splitting inhibition, flash inhibition and other means are introduced to process the tracks, so that the tracks of the vehicles are continuous and stable, the real-time vehicle motion situation is met, and the phenomena of serious influence on the appearance, such as track flash, track fracture, splitting ghost and the like, of the vehicles are avoided.
Referring to fig. 11, fig. 11 shows a block diagram of a track processing apparatus according to an embodiment of the application, an apparatus 1100 includes:
a matching module 1110, configured to perform association matching on a plurality of existing tracks before the target time point and a plurality of first targets observed at the target time point, so as to obtain a track association matching result;
the first determining module 1120 is configured to determine, if there are existing tracks of the plurality of existing tracks that are not associated with the first target according to the track association matching result, predict according to the existing tracks of the target, determine a prediction result corresponding to a second target associated with the existing track of the target before the target time at a prediction time point, where the prediction result includes a prediction position point and a prediction heading angle corresponding to the second target at the prediction time point;
a second determining module 1130, configured to determine a prediction confidence level of the prediction result according to a predicted heading angle corresponding to the predicted time point of the second target, a heading angle corresponding to the last position point in the existing track of the target, and the number of target predicted position points in the existing track of the target; the target predicted position point is a predicted position point positioned after the last observation position point in the existing track of the target;
And the association module 1140 is configured to perform track association on the predicted position point corresponding to the predicted time point of the second target and the existing track of the target if the predicted confidence coefficient is greater than or equal to the confidence coefficient threshold.
Optionally, the second determining module 1130 is further configured to calculate a difference between a predicted heading angle corresponding to the predicted time point of the second target and a heading angle corresponding to the last location point in the existing track of the target; and determining the prediction confidence of the prediction result according to the number and the difference of the target prediction position points.
Optionally, the plurality of first targets are observed by a plurality of sensors; the first determining module 1120 is further configured to determine, if there are intersections between the fields of view of the plurality of sensors at the target time point, and a target existing track that is not associated with the first target exists in the plurality of existing tracks according to the track association matching result, predict according to the target existing track, and determine a prediction result corresponding to a prediction time point of a second target associated with the target existing track before the target time point after the target time point.
Optionally, the plurality of first targets are observed by a plurality of sensors; the track association matching result comprises a first association result; a matching module 1110, configured to determine a candidate track from the plurality of existing tracks and determine a candidate target from the plurality of first targets according to the observation identifications of the plurality of sensors for each existing track and the observation identifications of the plurality of sensors for each first target; determining the approximation degree between the candidate track and the candidate target according to the observation identifications of the plurality of sensors for the candidate track and the observation identifications of the plurality of sensors for the candidate target; and carrying out track association on the candidate track and the candidate target according to the approximation degree to obtain a first association result.
Optionally, the matching module 1110 is further configured to construct, according to the observation identifiers of the plurality of sensors for each existing track, an observation identifier set corresponding to each existing track; if a plurality of observation identification sets exist, a target observation identification set comprising observation identifications of a plurality of sensors aiming at a first target is obtained, an existing track corresponding to the target observation identification set is taken as a candidate track, and the first target is obtained as a candidate target.
Optionally, the matching module 1110 is further configured to determine a ratio of observation identifiers of the plurality of sensors for the candidate targets in the set of observation identifiers corresponding to the candidate track, as an identifier ratio corresponding to the candidate target under the candidate track; and determining the approximation degree between the candidate track and the candidate target according to the mark duty ratio corresponding to the candidate target under the candidate track and the distance between the candidate track and the candidate target.
Optionally, the matching module 1110 is further configured to divide an identification duty ratio corresponding to the candidate target under the candidate track by a distance between the candidate track and the candidate target to obtain an approximation degree between the candidate track and the candidate target.
Optionally, the track association matching result further comprises a second association result; the matching module 1110 is further configured to determine, according to the first association result, that a first track that is not associated with the first object exists in the plurality of existing tracks and a third track that is not associated with the existing track exists in the plurality of first objects, and perform track association on the first track and the third target according to a distance between the first track and the third target, so as to obtain a second association result.
Optionally, the track association matching result further comprises a third association result; the matching module 1110 is further configured to, if it is determined, according to the first association result and the second association result, that there is a second track that is not associated with the first object in the plurality of existing tracks and there is a fourth object that is not associated with the existing track in the plurality of first objects, associate, for each fourth object, the fourth object to a second track that is closest to the fourth object, and obtain a third association result; and if the fifth target which is not associated with the existing track exists in the plurality of second targets according to the first association result, the second association result and the third association result, ignoring the fifth target.
Optionally, the plurality of first targets are observed at observation points; the matching module 1110 is further configured to determine, if the observation point is located in the target area of the intersection, according to the first association result and the second association result, that there is a second track that is not associated with the first target in the plurality of existing tracks and that there is a fourth target that is not associated with the existing track in the plurality of first targets, and for each fourth target, associate the fourth target to a second track that is closest to the fourth target in distance, thereby obtaining a third association result.
It should be noted that, in the present application, the device embodiment and the foregoing method embodiment correspond to each other, and specific principles in the device embodiment may refer to the content in the foregoing method embodiment, which is not described herein again.
Fig. 12 shows a block diagram of an electronic device for performing a trajectory processing method according to an embodiment of the present application. The electronic device may be the terminal 20 or the server 10 in fig. 1, and it should be noted that, the computer system 1200 of the electronic device shown in fig. 12 is only an example, and should not impose any limitation on the functions and the application scope of the embodiments of the present application.
As shown in fig. 12, the computer system 1200 includes a central processing unit (Central Processing Unit, CPU) 1201 which can perform various appropriate actions and processes, such as performing the methods in the above-described embodiments, according to a program stored in a Read-Only Memory (ROM) 1202 or a program loaded from a storage section 1208 into a random access Memory (Random Access Memory, RAM) 1203. In the RAM 1203, various programs and data required for the system operation are also stored. The CPU1201, ROM1202, and RAM 1203 are connected to each other through a bus 1204. An Input/Output (I/O) interface 1205 is also connected to bus 1204.
The following components are connected to the I/O interface 1205: an input section 1206 including a keyboard, a mouse, and the like; an output portion 1207 including a Cathode Ray Tube (CRT), a liquid crystal display (Liquid Crystal Display, LCD), and a speaker, etc.; a storage section 1208 including a hard disk or the like; and a communication section 1209 including a network interface card such as a LAN (Local Area Network ) card, a modem, or the like. The communication section 1209 performs communication processing via a network such as the internet. The drive 1210 is also connected to the I/O interface 1205 as needed. A removable medium 1211 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is installed on the drive 1210 as needed, so that a computer program read out therefrom is installed into the storage section 1208 as needed.
In particular, according to embodiments of the present application, the processes described above with reference to flowcharts may be implemented as computer software programs. For example, embodiments of the present application include a computer program product comprising a computer program embodied on a computer readable medium, the computer program comprising program code for performing the method shown in the flowcharts. In such an embodiment, the computer program can be downloaded and installed from a network via the communication portion 1209, and/or installed from the removable media 1211. When executed by a Central Processing Unit (CPU) 1201, performs the various functions defined in the system of the present application.
It should be noted that, the computer readable medium shown in the embodiments of the present application may be a computer readable signal medium or a computer readable storage medium, or any combination of the two. The computer readable storage medium can be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a combination of any of the foregoing. More specific examples of the computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-Only Memory (ROM), an erasable programmable read-Only Memory (Erasable Programmable Read Only Memory, EPROM), flash Memory, an optical fiber, a portable compact disc read-Only Memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the context of this document, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In the present application, however, the computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, with the computer-readable program code embodied therein. Such a propagated data signal may take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination of the foregoing. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wired, etc., or any suitable combination of the foregoing.
The flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present application. Where each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams or flowchart illustration, and combinations of blocks in the block diagrams or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
The units involved in the embodiments of the present application may be implemented by software, or may be implemented by hardware, and the described units may also be provided in a processor. Wherein the names of the units do not constitute a limitation of the units themselves in some cases.
As another aspect, the present application also provides a computer-readable storage medium that may be contained in the electronic device described in the above embodiment; or may exist alone without being incorporated into the electronic device. The computer readable storage medium carries computer readable instructions which, when executed by a processor, implement the method of any of the above embodiments.
According to an aspect of embodiments of the present application, there is provided a computer program product or computer program comprising computer instructions stored in a computer readable storage medium. The processor of the electronic device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions to cause the electronic device to perform the method of any of the embodiments described above.
It should be noted that although in the above detailed description several modules or units of a device for action execution are mentioned, such a division is not mandatory. Indeed, the features and functions of two or more modules or units described above may be embodied in one module or unit in accordance with embodiments of the application. Conversely, the features and functions of one module or unit described above may be further divided into a plurality of modules or units to be embodied.
From the above description of embodiments, those skilled in the art will readily appreciate that the example embodiments described herein may be implemented in software, or may be implemented in software in combination with the necessary hardware. Thus, the technical solution according to the embodiments of the present application may be embodied in the form of a software product, which may be stored in a non-volatile storage medium (may be a CD-ROM, a usb disk, a mobile hard disk, etc.) or on a network, and includes several instructions to cause an electronic device (may be a personal computer, a server, a touch terminal, or a network device, etc.) to perform the method according to the embodiments of the present application.
Other embodiments of the application will be apparent to those skilled in the art from consideration of the specification and practice of the embodiments disclosed herein. This application is intended to cover any variations, uses, or adaptations of the application following, in general, the principles of the application and including such departures from the present disclosure as come within known or customary practice within the art to which the application pertains. It is to be understood that the application is not limited to the precise arrangements and instrumentalities shown in the drawings, which have been described above, and that various modifications and changes may be effected without departing from the scope thereof. The scope of the application is limited only by the appended claims.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present application, and are not limiting; although the application has been described in detail with reference to the foregoing embodiments, it will be appreciated by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not drive the essence of the corresponding technical solutions to depart from the spirit and scope of the technical solutions of the embodiments of the present application.
Claims (13)
1. A track processing method, the method comprising:
performing association matching on a plurality of existing tracks before a target time point and a plurality of first targets observed at the target time point to obtain a track association matching result; the plurality of first targets is observed by a plurality of sensors;
if the intersection exists in the fields of view of the plurality of sensors at the target time point, determining that a target existing track which is not associated with a first target exists in the plurality of existing tracks according to the track association matching result, predicting according to the target existing track, and determining a prediction result corresponding to a prediction time point of a second target which is associated with the target existing track before the target time point after the target time point, wherein the prediction result comprises a prediction position point and a prediction course angle corresponding to the second target at the prediction time point;
Determining the prediction confidence of the prediction result according to the predicted course angle of the second target corresponding to the prediction time point, the course angle corresponding to the last position point in the existing target track and the number of target predicted position points in the existing target track; the target predicted position point is a predicted position point positioned behind the last observation position point in the existing track of the target;
and if the prediction confidence coefficient is greater than or equal to a confidence coefficient threshold value, carrying out track association on a predicted position point corresponding to the predicted time point of the second target and the existing track of the target.
2. The method of claim 1, wherein determining the prediction confidence of the prediction result according to the predicted heading angle of the second target corresponding to the predicted time point, the heading angle corresponding to the last position point in the existing track of the target, and the number of target predicted position points in the existing track of the target comprises:
calculating the difference between the predicted course angle corresponding to the predicted time point of the second target and the course angle corresponding to the last position point in the existing track of the target;
And determining the prediction confidence of the prediction result according to the number of the target prediction position points and the difference.
3. The method of claim 2, wherein the prediction confidence of the predicted outcome is inversely related to the number of target predicted location points and the prediction confidence of the predicted outcome is inversely related to the difference.
4. The method of claim 1, wherein the trajectory-association matching result comprises a first association result;
performing association matching on a plurality of existing tracks before a target time point and a plurality of first targets observed at the target time point to obtain a track association matching result, wherein the method comprises the following steps:
determining a candidate track from the plurality of existing tracks and a candidate target from the plurality of first targets according to the observation identifications of the plurality of sensors for each of the existing tracks and the observation identifications of the plurality of sensors for each of the first targets;
determining the approximation degree between the candidate track and the candidate target according to the observation identifications of the plurality of sensors for the candidate track and the observation identifications of the plurality of sensors for the candidate target;
And carrying out track association on the candidate track and the candidate target according to the approximation degree to obtain the first association result.
5. The method of claim 4, wherein the determining a candidate track from the plurality of existing tracks and a candidate target from the plurality of first targets based on the observed identity of the plurality of sensors for each of the existing tracks and the observed identity of the plurality of sensors for each of the first targets comprises:
constructing an observation identification set corresponding to each existing track according to the observation identifications of the plurality of sensors aiming at each existing track;
if a target observation identification set comprising the observation identifications of the plurality of sensors aiming at the first target exists in the plurality of observation identification sets, acquiring an existing track corresponding to the target observation identification set as a candidate track, and acquiring the first target as a candidate target.
6. The method of claim 4, wherein the determining the proximity between the candidate trajectory and the candidate target based on the observed identifications of the plurality of sensors for the candidate trajectory and the observed identifications of the plurality of sensors for the candidate target comprises:
Determining the duty ratio of the observation identifications of the plurality of sensors aiming at the candidate targets in the observation identification set corresponding to the candidate track as the identification duty ratio corresponding to the candidate targets under the candidate track;
and determining the approximation degree between the candidate track and the candidate target according to the mark duty ratio corresponding to the candidate target under the candidate track and the distance between the candidate track and the candidate target.
7. The method of claim 6, wherein determining the proximity between the candidate trajectory and the candidate target based on the identification duty cycle corresponding to the candidate target under the candidate trajectory and the distance between the candidate trajectory and the candidate target comprises:
dividing the mark duty ratio corresponding to the candidate target under the candidate track with the distance between the candidate track and the candidate target to obtain the approximation degree between the candidate track and the candidate target.
8. The method of claim 4, wherein the trajectory correlation match result further comprises a second correlation result;
performing association matching on a plurality of existing tracks before a target time point and a plurality of first targets observed at the target time point to obtain a track association matching result, and further comprising:
And if the first track which is not associated with the first target exists in the plurality of existing tracks and the third target which is not associated with the existing track exists in the plurality of first targets according to the first association result, carrying out track association on the first track and the third target according to the distance between the first track and the third target, and obtaining a second association result.
9. The method of claim 8, wherein the trajectory correlation match result further comprises a third correlation result;
performing association matching on a plurality of existing tracks before a target time point and a plurality of first targets observed at the target time point to obtain a track association matching result, and further comprising:
if a second track which is not associated with the first track exists in the plurality of existing tracks and a fourth track which is not associated with the existing track exists in the plurality of first tracks according to the first association result and the second association result, associating the fourth track with a second track closest to the fourth track for each fourth target, and obtaining a third association result;
and if the fifth target which is not associated with the existing track exists in the second targets according to the first association result, the second association result and the third association result, ignoring the fifth target.
10. The method of claim 9, wherein the plurality of first targets are observed at an observation point;
if, according to the first association result and the second association result, a second track which is not associated with the first track exists in the plurality of existing tracks and a fourth track which is not associated with the existing track exists in the plurality of first targets, for each fourth target, the fourth target is associated to a second track which is closest to the fourth target, and a third association result is obtained, including:
and if the measuring point observation is positioned in the target area of the intersection, determining that a second track which is not associated with the first target exists in the plurality of existing tracks and a fourth target which is not associated with the existing track exists in the plurality of first targets according to the first association result and the second association result, and associating the fourth target to a second track which is closest to the fourth target in distance aiming at each fourth target to obtain a third association result.
11. A track processing device, the device comprising:
the matching module is used for carrying out association matching on a plurality of existing tracks before a target time point and a plurality of first targets observed at the target time point to obtain track association matching results; the plurality of first targets is observed by a plurality of sensors;
The first determining module is used for determining that a target existing track which is not associated with a first target exists in the plurality of existing tracks according to the track association matching result if intersection exists among the fields of view of the plurality of sensors at the target time point, predicting according to the target existing track, and determining a prediction result corresponding to a prediction time point of a second target which is associated with the target existing track before the target time point after the target time point, wherein the prediction result comprises a prediction position point and a prediction course angle corresponding to the prediction time point of the second target;
the second determining module is used for determining the prediction confidence of the prediction result according to the predicted course angle of the second target corresponding to the prediction time point, the course angle corresponding to the last position point in the existing target track and the number of target predicted position points in the existing target track; the target predicted position point is a predicted position point positioned behind the last observation position point in the existing track of the target;
and the association module is used for carrying out track association on the predicted position point corresponding to the predicted time point of the second target and the existing track of the target if the predicted confidence coefficient is greater than or equal to a confidence coefficient threshold value.
12. An electronic device, comprising:
one or more processors;
a memory;
one or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, the one or more applications configured to perform the method of any of claims 1-10.
13. A computer readable storage medium, characterized in that the computer readable storage medium has stored therein a program code, which is callable by a processor for performing the method according to any one of claims 1-10.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310999360.6A CN116718197B (en) | 2023-08-09 | 2023-08-09 | Track processing method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310999360.6A CN116718197B (en) | 2023-08-09 | 2023-08-09 | Track processing method and device, electronic equipment and storage medium |
Publications (2)
Publication Number | Publication Date |
---|---|
CN116718197A CN116718197A (en) | 2023-09-08 |
CN116718197B true CN116718197B (en) | 2023-10-24 |
Family
ID=87873812
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310999360.6A Active CN116718197B (en) | 2023-08-09 | 2023-08-09 | Track processing method and device, electronic equipment and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116718197B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117672007B (en) * | 2024-02-03 | 2024-04-26 | 福建省高速公路科技创新研究院有限公司 | Road construction area safety precaution system based on thunder fuses |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2017185688A1 (en) * | 2016-04-26 | 2017-11-02 | 深圳大学 | Method and apparatus for tracking on-line target |
CN107423686A (en) * | 2017-06-15 | 2017-12-01 | 深圳大学 | Video multi-target Fuzzy data association method and device |
WO2018107492A1 (en) * | 2016-12-16 | 2018-06-21 | 深圳大学 | Intuitionistic fuzzy random forest-based method and device for target tracking |
WO2018227491A1 (en) * | 2017-06-15 | 2018-12-20 | 深圳大学 | Method and device for association of fuzzy data of multiple targets in video |
CN113291320A (en) * | 2021-06-15 | 2021-08-24 | 苏州智加科技有限公司 | Vehicle track prediction method, device, equipment and storage medium |
CN114663471A (en) * | 2022-02-17 | 2022-06-24 | 深圳大学 | Target tracking method and device and computer readable storage medium |
CN116228989A (en) * | 2023-03-30 | 2023-06-06 | 北京数原数字化城市研究中心 | Three-dimensional track prediction method, device, equipment and medium |
CN116266360A (en) * | 2021-12-16 | 2023-06-20 | 长安大学 | Vehicle target detection tracking method based on multi-source information fusion |
CN116331259A (en) * | 2023-03-16 | 2023-06-27 | 同济大学 | Vehicle multi-mode track prediction method based on semi-supervised model |
-
2023
- 2023-08-09 CN CN202310999360.6A patent/CN116718197B/en active Active
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2017185688A1 (en) * | 2016-04-26 | 2017-11-02 | 深圳大学 | Method and apparatus for tracking on-line target |
WO2018107492A1 (en) * | 2016-12-16 | 2018-06-21 | 深圳大学 | Intuitionistic fuzzy random forest-based method and device for target tracking |
CN107423686A (en) * | 2017-06-15 | 2017-12-01 | 深圳大学 | Video multi-target Fuzzy data association method and device |
WO2018227491A1 (en) * | 2017-06-15 | 2018-12-20 | 深圳大学 | Method and device for association of fuzzy data of multiple targets in video |
CN113291320A (en) * | 2021-06-15 | 2021-08-24 | 苏州智加科技有限公司 | Vehicle track prediction method, device, equipment and storage medium |
CN116266360A (en) * | 2021-12-16 | 2023-06-20 | 长安大学 | Vehicle target detection tracking method based on multi-source information fusion |
CN114663471A (en) * | 2022-02-17 | 2022-06-24 | 深圳大学 | Target tracking method and device and computer readable storage medium |
CN116331259A (en) * | 2023-03-16 | 2023-06-27 | 同济大学 | Vehicle multi-mode track prediction method based on semi-supervised model |
CN116228989A (en) * | 2023-03-30 | 2023-06-06 | 北京数原数字化城市研究中心 | Three-dimensional track prediction method, device, equipment and medium |
Non-Patent Citations (1)
Title |
---|
基于相机雷达融合的改进GM-PHD多目标跟踪算法;张晗等;;传感器与微***(第05期);123-127 * |
Also Published As
Publication number | Publication date |
---|---|
CN116718197A (en) | 2023-09-08 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN113221677B (en) | Track abnormality detection method and device, road side equipment and cloud control platform | |
CN113264066B (en) | Obstacle track prediction method and device, automatic driving vehicle and road side equipment | |
CN116718197B (en) | Track processing method and device, electronic equipment and storage medium | |
CN113859264B (en) | Vehicle control method, device, electronic equipment and storage medium | |
CN112528786A (en) | Vehicle tracking method and device and electronic equipment | |
CN112528927B (en) | Confidence determining method based on track analysis, road side equipment and cloud control platform | |
CN114663471A (en) | Target tracking method and device and computer readable storage medium | |
CN114842445A (en) | Target detection method, device, equipment and medium based on multi-path fusion | |
CN114179832A (en) | Lane changing method for autonomous vehicle | |
CN113177980B (en) | Target object speed determining method and device for automatic driving and electronic equipment | |
CN114528941A (en) | Sensor data fusion method and device, electronic equipment and storage medium | |
CN113763425A (en) | Road area calibration method and electronic equipment | |
CN114018269A (en) | Positioning method, positioning device, electronic equipment, storage medium and automatic driving vehicle | |
CN113887391A (en) | Method and device for recognizing road sign and automatic driving vehicle | |
CN116563802A (en) | Tunnel scene judging method, device, equipment and storage medium | |
CN115953434B (en) | Track matching method, track matching device, electronic equipment and storage medium | |
CN115546597A (en) | Sensor fusion method, device, equipment and storage medium | |
CN115330841A (en) | Method, apparatus, device and medium for detecting projectile based on radar map | |
CN116996760A (en) | Video data processing method and device, computer readable medium and electronic equipment | |
Xu et al. | An efficient multi‐sensor fusion and tracking protocol in a vehicle‐road collaborative system | |
CN112507957A (en) | Vehicle association method and device, road side equipment and cloud control platform | |
Ding et al. | Who is partner: A new perspective on data association of multi-object tracking | |
CN117962930B (en) | Unmanned vehicle control method and device, unmanned vehicle and computer readable storage medium | |
CN114049615B (en) | Traffic object fusion association method and device in driving environment and edge computing equipment | |
CN116883707A (en) | Millimeter wave radar point cloud clustering method |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
REG | Reference to a national code |
Ref country code: HK Ref legal event code: DE Ref document number: 40093146 Country of ref document: HK |