CN109765886A - The target trajectory recognition methods that vehicle follows - Google Patents
The target trajectory recognition methods that vehicle follows Download PDFInfo
- Publication number
- CN109765886A CN109765886A CN201811572884.2A CN201811572884A CN109765886A CN 109765886 A CN109765886 A CN 109765886A CN 201811572884 A CN201811572884 A CN 201811572884A CN 109765886 A CN109765886 A CN 109765886A
- Authority
- CN
- China
- Prior art keywords
- target
- information
- vehicle
- image information
- following target
- 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.)
- Granted
Links
Landscapes
- Traffic Control Systems (AREA)
Abstract
The present invention provides a kind of target trajectory recognition methods that vehicle follows, comprising: obtains the location information of vehicle;According to the location information of the vehicle, map datum is obtained;In preset duration, the multiple image information for following target that vehicle is followed is obtained;Every frame described image information includes temporal information when obtaining described image information;According to the temporal information, processing is fitted to the multiple image information and the map datum, obtains the location information for following target corresponding with each temporal information;The multiple location informations for following target are spliced, the target trajectory for following target is obtained.The time for obtaining track is saved as a result, improves the efficiency of track acquisition.
Description
Technical field
The present invention relates to image and data processing field more particularly to a kind of target trajectory recognition methods that vehicle follows.
Background technique
With the development of economy and the emergence of artificial intelligence technology, automatic Pilot technology is also increasingly by the blueness in market
It looks at.By automatic Pilot technical application in cleaning applications, automatic cleaning equipment is generated.
In the prior art, automatic cleaning equipment passes through itself multiple sensor combinations, by the algorithm of a series of complex,
Obtain the track of the motive objects in front.The defects such as that there is operands is big for this mode, computational efficiency is low.
Summary of the invention
The purpose of the embodiment of the present invention is that a kind of target trajectory recognition methods that vehicle follows is provided, to solve the prior art
The problem of.
To solve the above problems, the present invention provides a kind of target trajectory recognition methods that vehicle follows, the method packet
It includes:
Obtain the location information of vehicle;
According to the location information of the vehicle, map datum is obtained;
In preset duration, the multiple image information for following target that vehicle is followed is obtained;Every frame described image information
Including obtaining temporal information when described image information;
According to the temporal information, processing is fitted to the multiple image information and the map datum, obtain and
The corresponding location information for following target of each temporal information;
The multiple location informations for following target are spliced, the target trajectory for following target is obtained.
Preferably, described in preset duration, before obtaining the multiple image information for following target that vehicle is followed, institute
State method further include:
By the video information for following target described in the acquisition of the first acquisition device;
The video information is handled, the image information for following target is obtained.
Preferably, the method also includes:
By the environment sensing data for following target described in the acquisition of the second acquisition device;
The environment sensing data are handled, laser point cloud data is generated;
According to the laser point cloud data, described image information is modified.
Preferably, described according to the temporal information, the multiple image information and the map datum are fitted
Processing obtains the location information for following target corresponding with each temporal information, specifically includes:
Every frame described image information is handled, the environmental data in every frame described image information is obtained;
The environmental data and the map datum are fitted, when the environmental data and the map datum are overlapped
When, determine position of the environmental data in the map datum;
According to the position, the corresponding location information for following target of every frame image is determined.
Preferably, described that the multiple location informations for following target are spliced, obtain the mesh for following target
Track is marked, is specifically included:
According to the temporal information, each location information for following target is ranked up;
According to ranking results, spliced, obtains the target trajectory for following target.
Preferably, the method also includes:
It obtains the vehicle and follows the distance between target with described;
When the distance is not less than preset distance threshold, the first warning message is generated;
By the audio playing unit of vehicle, first warning message is played.
Preferably, after the method further include:
When the distance is not less than preset distance threshold, the second warning message is generated;The second warning message packet
Include the estimated waiting time for following target;
By second warning message be sent to it is described follow target so that described follow target estimated to be waited according to described
Duration is waited.
Preferably, the previous moment not less than preset distance threshold is obtained respectively and is not less than preset distance threshold
When, follow the image information of target;
Image when to the previous moment not less than preset distance threshold and not less than preset distance threshold
Information is handled, and determines the velocity information for following target;
According to the velocity information for following target, the velocity information of vehicle and safe distance, when calculating estimated wait
It is long.
Preferably, when the first acquisition device is binocular camera, parameter and every frame institute by the binocular camera
Image information is stated, the vehicle is calculated with described and follows the distance between target.
Preferably, the ultrasound data acquired by third acquisition device, calculate the vehicle and it is described follow target it
Between distance.
By the target trajectory recognition methods followed using vehicle provided by the invention, the location information of vehicle is obtained;Root
According to the location information of the vehicle, map datum is obtained;In preset duration, the multiframe for following target that vehicle is followed is obtained
Image information;Every frame described image information includes temporal information when obtaining described image information;It is right according to the temporal information
The multiple image information and the map datum are fitted processing, obtain it is corresponding with each temporal information it is described with
With the location information of target;The multiple location informations for following target are spliced, the target for following target is obtained
Track.The time for obtaining track is saved as a result, improves the efficiency of track acquisition.
Detailed description of the invention
Fig. 1 is the target trajectory recognition methods flow diagram that vehicle provided in an embodiment of the present invention follows.
Specific embodiment
The application is described in further detail with reference to the accompanying drawings and examples.It is understood that this place is retouched
The specific embodiment stated is used only for explaining related invention, rather than the restriction to the invention.It also should be noted that for just
Part relevant to related invention is illustrated only in description, attached drawing.
It should be noted that in the absence of conflict, the features in the embodiments and the embodiments of the present application can phase
Mutually combination.The application is described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.
It before application method provided by the invention, first to determine and follow target, follow target on how to determination, it can be with
By the method for image feature comparison, for example, when applying the method in suspicion tracking, in the memory of vehicle, storage
Have suspicion license plate number, the facial characteristics of suspicion personnel, suspicion personnel image information, vehicle is by acquired image information and pre-
The license plate number or facial characteristics deposited are matched, when matching degree is greater than some numerical value, then through server Secondary Match, with into
Row confirmation, when being confirmed to be suspicion object, follows it.It, can be in vehicle when applying the method in cleaning applications
Memory in be stored with the feature of cleanup crew, or the terminal in advance carrying cleanup crew ID and vehicle ID into
Row binding, after server interaction, determination follows target.
Fig. 1 is the target trajectory recognition methods flow diagram that vehicle provided in an embodiment of the present invention follows.This method is answered
In automatic driving vehicle, especially in the cleaning vehicle of automatic Pilot, the available target trajectory for following target of vehicle is cleaned,
And it is performed corresponding processing according to the target trajectory, for example follow.The executing subject of this method can be automatic driving vehicle
Control unit.Control unit for vehicle can be understood as the control module for controlling vehicle driving.As shown in Figure 1, this method
The following steps are included:
Step 101, the location information of vehicle is obtained.
Specifically, the locating module on vehicle, such as global positioning system (Global Positioning can be passed through
System, GPS) obtain vehicle itself location information.It can also be by sending query messages, resolution server hair to server
After the response message of the carrying location information sent, the location information of vehicle itself is obtained.
Step 102, according to the location information of vehicle, map datum is obtained.
Specifically, when vehicle is in a certain position the position can be loaded according to the location information of location
Map, for example, vehicle is in the street A, the longitude and latitude data of the location information of vehicle are known at this time, can be by the street A
Upper level unit, the map in the city A loaded, and can also be loaded longitude and latitude according to the longitude and latitude data and is in a certain range
When corresponding map datum.As to how load, can be and download from server, be also possible to vehicle and load in advance, this Shen
Please this is not limited.
Step 103, in preset duration, the multiple image information for following target that vehicle is followed is obtained.
Wherein, according to the difference for following scene, follow target also different.It, should when applying the method in cleaning applications
Target is followed to can be cleaner or cleaning vehicle, when applying the method in suspect's tracking, this follows target can be
Suspicion personnel or suspected vehicles.
Preset duration can be 10 minutes, in 10 minutes, get multiple image information.Every frame image information includes
Obtain temporal information when image information.
Specifically, there is the first acquisition device, to acquire the video information for following target on vehicle.The first acquisition dress
It sets and can be camera, subsequent in order to calculate vehicle and follow the distance between target, camera can be binocular camera.
Binocular camera can be handled the video data of acquisition, video frame be extracted, to extract from video frame
Image information out.
Further, on vehicle, other than the first acquisition device, there is also other acquisition devices, such as the second acquisition dress
It sets, the second acquisition device can be laser radar.Image information can be corrected by the data of laser radar.
Specifically, handling environment sensing data, laser point cloud data is generated;
According to laser point cloud data, image information is modified.
Wherein, amendment herein, also can be regarded as fusion treatment.Fusion treatment, for example can be and enhanced using details
Algorithm carries out details enhancing.
Step 104, according to temporal information, processing is fitted to multiple image information and map datum, is obtained and each
The corresponding location information for following target of temporal information.
It, can be by acquisition when cannot be with vehicle interaction locations information specifically, when following target for some reason
To image information handled, to obtain the location information for following target.Location information can be obtained through the following steps.
Firstly, handling every frame image information, the environmental data in every frame image information is obtained.
Then, environmental data and map datum are fitted, when environmental data and map datum coincidence, determine environment
Position of the data in map datum.
Finally, determining the corresponding location information for following target of every frame image according to position.
It wherein, include environmental data, such as building mark, traffic mark, road markings etc. in image information.
After environmental data and map datum are fitted, the same characteristic features in the two, such as barrier can be carried out
Integrated treatment calculates the location information for following target.
Obstacle information herein can be building, fixed traffic sign (ratio on fixed obstacle, such as map
Such as, for fixing the bar of traffic lights), fixed object (for example, static vehicle, pedestrian, curb).These obstacle informations, can
To directly obtain by image information and map datum.
Step 105, the multiple location informations for following target are spliced, obtains the target trajectory for following target.
Specifically, can be ranked up according to temporal information to each location information for following target;It is tied according to sequence
Fruit is spliced, and the target trajectory for following target is obtained.
For example, collected location information include 1,2,3,4,5, corresponding temporal information be 10:51,10:52,10:54,
10:53 and 10:55 is then ordered as 1,2,4,3,5 to these location informations, splices to it, available target trajectory.
Further, it before generating target trajectory, in order to improve the accuracy of location information, can also be adopted using second
The environment sensing data of acquisition means are modified the determining location information for following target.
Specifically, sensing module can be the laser radar installed on vehicle, and in vehicle travel process, the obstacle of surrounding
Object information, such as lane line, the barrier of movement etc., the traffic lights of variation, in conjunction with above-mentioned obstacle information and herein
The obstacle information perceived in driving process, after fusion treatment, available final obstacle information, referred to as target obstacle
Information.
According to target obstacle information, in the track of the splicing and track of target obstacle coincidence is modified, thus
Generate target trajectory.
Further, after step 105, further includes: obtain vehicle and follow the distance between target;
When distance is not less than preset distance threshold, the first warning message is generated;
By the audio playing unit of vehicle, the first warning message is played.
Specifically, when vehicle and follow it is excessive at a distance from target, be more than distance threshold when, calculating distance after, generation
Signal is controlled, which can control audio playing unit and play the first warning message.First warning message can be language
Speech casting, is also possible to certain specific sound.
In one example, when the first acquisition device is binocular camera, parameter and every frame by binocular camera
Image information calculates vehicle and follows the distance between target.Wherein it is possible to using the binocular range measurement principle of binocular camera,
It calculates vehicle and follows the distance between target.
In another example, the ultrasound data acquired by third acquisition device, calculate vehicle and follow target it
Between distance.Wherein, third acquisition device is ultrasonic radar, can use supersonic sounding, calculate vehicle and follow target it
Between distance.
Further, after method further include:
When distance is not less than preset distance threshold, the second warning message is generated;Second warning message includes following mesh
Target estimated waiting time;
Second warning message is sent to and follows target, waiting time is waited on the estimation so as to follow target.
Specifically, when following target to clean vehicle or cleaner, if target is followed to exceed at a distance from vehicle
The second warning message can be generated in distance threshold, vehicle, and the second warning message may include estimated waiting time.
Wherein, vehicle can calculate according to the image information acquired beyond distance threshold and previous moment and follow mesh
Target location information and velocity information, and safe distance when according to velocity information, itself velocity information and vehicle driving,
Calculate the estimated waiting time for following target.
By the target trajectory recognition methods followed using vehicle provided in an embodiment of the present invention, it can directly pass through image
Information gets the track for following target, saves the time for obtaining track, improves the efficiency of track acquisition.
Professional should further appreciate that, described in conjunction with the examples disclosed in the embodiments of the present disclosure
Unit and algorithm steps, can be realized with electronic hardware, computer software, or a combination of the two, hard in order to clearly demonstrate
The interchangeability of part and software generally describes each exemplary composition and step according to function in the above description.
These functions are implemented in hardware or software actually, the specific application and design constraint depending on technical solution.
Professional technician can use different methods to achieve the described function each specific application, but this realization
It should not be considered as beyond the scope of the present invention.
The step of method described in conjunction with the examples disclosed in this document or algorithm, can be executed with hardware, processor
The combination of software module or the two is implemented.Software module can be placed in random access memory (RAM), memory, read-only memory
(ROM), electrically programmable ROM, electrically erasable ROM, register, hard disk, moveable magnetic disc, CD-ROM or technical field
In any other form of storage medium well known to interior.
Above specific embodiment has carried out further in detail the purpose of the present invention, technical scheme and beneficial effects
Illustrate, it should be understood that the above is only a specific embodiment of the invention, the protection model that is not intended to limit the present invention
It encloses, all within the spirits and principles of the present invention, any modification, equivalent substitution, improvement and etc. done should be included in the present invention
Protection scope within.
Claims (10)
1. a kind of target trajectory recognition methods that vehicle follows, which is characterized in that the described method includes:
Obtain the location information of vehicle;
According to the location information of the vehicle, map datum is obtained;
In preset duration, the multiple image information for following target that vehicle is followed is obtained;Every frame described image information includes
Obtain temporal information when described image information;
According to the temporal information, processing is fitted to the multiple image information and the map datum, is obtained and each
The corresponding location information for following target of the temporal information;
The multiple location informations for following target are spliced, the target trajectory for following target is obtained.
2. the method according to claim 1, wherein described in preset duration, obtain that vehicle followed with
Before the multiple image information of target, the method also includes:
By the video information for following target described in the acquisition of the first acquisition device;
The video information is handled, the image information for following target is obtained.
3. according to the method described in claim 2, it is characterized in that, the method also includes:
By the environment sensing data for following target described in the acquisition of the second acquisition device;
The environment sensing data are handled, laser point cloud data is generated;
According to the laser point cloud data, described image information is modified.
4. the method according to claim 1, wherein described according to the temporal information, to the multiple image
Information and the map datum are fitted processing, obtain the position for following target corresponding with each temporal information
Information specifically includes:
Every frame described image information is handled, the environmental data in every frame described image information is obtained;
The environmental data and the map datum are fitted, when the environmental data and map datum coincidence,
Determine position of the environmental data in the map datum;
According to the position, the corresponding location information for following target of every frame image is determined.
5. the method according to claim 1, wherein described carry out the multiple location informations for following target
Splicing obtains the target trajectory for following target, specifically includes:
According to the temporal information, each location information for following target is ranked up;
According to ranking results, spliced, obtains the target trajectory for following target.
6. the method according to claim 1, wherein the method also includes:
It obtains the vehicle and follows the distance between target with described;
When the distance is not less than preset distance threshold, the first warning message is generated;
By the audio playing unit of vehicle, first warning message is played.
7. the method according to claim 1, wherein after the method further include:
When the distance is not less than preset distance threshold, the second warning message is generated;Second warning message include with
With the estimated waiting time of target;
By second warning message be sent to it is described follow target so that described follow target according to the estimated waiting time
It is waited.
8. the method according to the description of claim 7 is characterized in that respectively obtain not less than preset distance threshold it is previous when
When carving and being not less than preset distance threshold, the image information of target is followed;
Image information when to the previous moment not less than preset distance threshold and not less than preset distance threshold
It is handled, determines the velocity information for following target;
According to the velocity information for following target, the velocity information of vehicle and safe distance, estimated waiting time is calculated.
9. according to method described in claim 6-8 any one, which is characterized in that when the first acquisition device is binocular camera
When, by the parameter and every frame described image information of the binocular camera, calculates the vehicle and described follow between target
Distance.
10. according to method described in claim 6-8 any one, which is characterized in that acquired by third acquisition device super
Sonic data calculates the vehicle with described and follows the distance between target.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811572884.2A CN109765886B (en) | 2018-12-21 | 2018-12-21 | Target track identification method followed by vehicle |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811572884.2A CN109765886B (en) | 2018-12-21 | 2018-12-21 | Target track identification method followed by vehicle |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109765886A true CN109765886A (en) | 2019-05-17 |
CN109765886B CN109765886B (en) | 2022-05-24 |
Family
ID=66450857
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811572884.2A Active CN109765886B (en) | 2018-12-21 | 2018-12-21 | Target track identification method followed by vehicle |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109765886B (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111368260A (en) * | 2020-03-18 | 2020-07-03 | 东软睿驰汽车技术(上海)有限公司 | Vehicle following method and device |
CN112810625A (en) * | 2021-04-19 | 2021-05-18 | 北京三快在线科技有限公司 | Method and device for correcting track |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101769745A (en) * | 2009-01-07 | 2010-07-07 | 宏达国际电子股份有限公司 | Mobile target tracking and navigation method, device thereof and computer program product used |
CN105015547A (en) * | 2014-04-28 | 2015-11-04 | 丰田自动车株式会社 | Driving assistance apparatus |
CN106289295A (en) * | 2016-08-30 | 2017-01-04 | 深圳市轱辘车联数据技术有限公司 | The vehicle follower method of a kind of self-driving travel and device |
CN106384540A (en) * | 2016-10-20 | 2017-02-08 | 深圳市元征科技股份有限公司 | Vehicle real-time track prediction method and prediction system |
US20170132931A1 (en) * | 1998-01-27 | 2017-05-11 | Blanding Hovenweep, Llc | Detection and alert of automobile braking event |
CN108362293A (en) * | 2018-02-24 | 2018-08-03 | 中电福富信息科技有限公司 | A kind of track of vehicle matching process based on key point technology |
US20180321682A1 (en) * | 2015-05-07 | 2018-11-08 | Kpit Technologies Limited | Guidance control system for autonomous-traveling vehicle |
-
2018
- 2018-12-21 CN CN201811572884.2A patent/CN109765886B/en active Active
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20170132931A1 (en) * | 1998-01-27 | 2017-05-11 | Blanding Hovenweep, Llc | Detection and alert of automobile braking event |
CN101769745A (en) * | 2009-01-07 | 2010-07-07 | 宏达国际电子股份有限公司 | Mobile target tracking and navigation method, device thereof and computer program product used |
CN105015547A (en) * | 2014-04-28 | 2015-11-04 | 丰田自动车株式会社 | Driving assistance apparatus |
US20180321682A1 (en) * | 2015-05-07 | 2018-11-08 | Kpit Technologies Limited | Guidance control system for autonomous-traveling vehicle |
CN106289295A (en) * | 2016-08-30 | 2017-01-04 | 深圳市轱辘车联数据技术有限公司 | The vehicle follower method of a kind of self-driving travel and device |
CN106384540A (en) * | 2016-10-20 | 2017-02-08 | 深圳市元征科技股份有限公司 | Vehicle real-time track prediction method and prediction system |
CN108362293A (en) * | 2018-02-24 | 2018-08-03 | 中电福富信息科技有限公司 | A kind of track of vehicle matching process based on key point technology |
Non-Patent Citations (2)
Title |
---|
WANG YAJING等: "Vector Field Based Sliding Mode Control of Curved Path Following for Miniature Unmanned Aerial Vehicles in Winds", 《JOURNAL OF SYSTEMS SCIENCE & COMPLEXITY》 * |
曹永军等: "多源交通信息采集、处理与发布***研究", 《自动化与信息工程》 * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111368260A (en) * | 2020-03-18 | 2020-07-03 | 东软睿驰汽车技术(上海)有限公司 | Vehicle following method and device |
CN111368260B (en) * | 2020-03-18 | 2023-09-05 | 东软睿驰汽车技术(上海)有限公司 | Vehicle following method and device |
CN112810625A (en) * | 2021-04-19 | 2021-05-18 | 北京三快在线科技有限公司 | Method and device for correcting track |
Also Published As
Publication number | Publication date |
---|---|
CN109765886B (en) | 2022-05-24 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107527092B (en) | Training algorithms for collision avoidance using auditory data | |
US11487988B2 (en) | Augmenting real sensor recordings with simulated sensor data | |
US10031526B1 (en) | Vision-based driving scenario generator for autonomous driving simulation | |
US10474964B2 (en) | Training algorithm for collision avoidance | |
CN111351493A (en) | Positioning method and system | |
CN108021862A (en) | Road sign identifies | |
GB2547999A (en) | Tracking objects within a dynamic environment for improved localization | |
GB2560805A (en) | Training an automatic traffic light detection model using simulated images | |
CN109686031A (en) | Identification follower method based on security protection | |
CN108399752A (en) | A kind of driving infractions pre-judging method, device, server and medium | |
CN109682388A (en) | Follow the determination method in path | |
JP2019039831A (en) | Automatic driving device | |
JP7194130B2 (en) | A method and apparatus for detecting emergency vehicles in real time and planning driving routes to deal with situations expected to be caused by emergency vehicles. | |
CN113640822B (en) | High-precision map construction method based on non-map element filtering | |
US11477567B2 (en) | Method and system for locating an acoustic source relative to a vehicle | |
CN109653140A (en) | Vehicle follower method in road cleaning operation | |
CN109741367A (en) | The target trajectory recognition methods that vehicle follows | |
CN109765886A (en) | The target trajectory recognition methods that vehicle follows | |
CN110696826A (en) | Method and device for controlling a vehicle | |
CN115273027A (en) | Environment sensing method, domain controller, storage medium and vehicle | |
CN109709953A (en) | Vehicle follower method in road cleaning operation | |
CN106240454B (en) | System for providing vehicle collision early warning and vehicle-mounted equipment | |
CN114333339B (en) | Deep neural network functional module de-duplication method | |
JP2018180641A (en) | Vehicle identification device | |
CN109633719A (en) | The target trajectory recognition methods that vehicle follows |
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 | ||
CP01 | Change in the name or title of a patent holder | ||
CP01 | Change in the name or title of a patent holder |
Address after: B4-006, maker Plaza, 338 East Street, Huilongguan town, Changping District, Beijing 100096 Patentee after: Beijing Idriverplus Technology Co.,Ltd. Address before: B4-006, maker Plaza, 338 East Street, Huilongguan town, Changping District, Beijing 100096 Patentee before: Beijing Idriverplus Technology Co.,Ltd. |