CN113706870B - Method for collecting main vehicle lane change data in congested scene and related equipment - Google Patents
Method for collecting main vehicle lane change data in congested scene and related equipment Download PDFInfo
- Publication number
- CN113706870B CN113706870B CN202111007467.5A CN202111007467A CN113706870B CN 113706870 B CN113706870 B CN 113706870B CN 202111007467 A CN202111007467 A CN 202111007467A CN 113706870 B CN113706870 B CN 113706870B
- Authority
- CN
- China
- Prior art keywords
- data
- data frame
- time
- lane change
- scene
- 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
- 230000008859 change Effects 0.000 title claims abstract description 91
- 238000000034 method Methods 0.000 title claims abstract description 76
- 238000012216 screening Methods 0.000 claims abstract description 29
- 230000004931 aggregating effect Effects 0.000 claims abstract description 18
- 230000008569 process Effects 0.000 claims description 39
- 238000013480 data collection Methods 0.000 claims description 10
- 230000002776 aggregation Effects 0.000 claims description 5
- 238000004220 aggregation Methods 0.000 claims description 5
- 238000004590 computer program Methods 0.000 claims description 4
- 238000007781 pre-processing Methods 0.000 claims description 4
- 230000000903 blocking effect Effects 0.000 claims 1
- 238000000638 solvent extraction Methods 0.000 claims 1
- 238000005516 engineering process Methods 0.000 abstract description 7
- 238000011161 development Methods 0.000 abstract description 5
- 238000004891 communication Methods 0.000 description 7
- 238000004422 calculation algorithm Methods 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 4
- 238000012545 processing Methods 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 238000001914 filtration Methods 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/16—Anti-collision systems
- G08G1/167—Driving aids for lane monitoring, lane changing, e.g. blind spot detection
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Traffic Control Systems (AREA)
Abstract
The application discloses a method for collecting main vehicle lane change data in a congested scene and related equipment, which comprises the following steps: acquiring automatic driving data of a main vehicle, wherein the automatic driving data comprises data under various themes; dividing the automatic driving data from a time dimension to obtain each data frame taking time as an index, wherein the data frame comprises data under each theme in the belonged time; combining semantic map data, and screening out target data frames of the main vehicle with lane change in a congested scene from all data frames; and aggregating the target data frames according to preset conditions to obtain lane change data in a congested scene. According to the method and the device, data of lane changing of the main vehicle in a congestion scene can be collected, and related data support is provided for development of an automatic driving technology.
Description
Technical Field
The application relates to the technical field of unmanned driving, in particular to a method for collecting main vehicle lane change data in a congested scene and related equipment.
Background
The automatic driving system can not well process complex scenes and can directly reflect the quality of the automatic driving technology. The lane change of the main vehicle in the congested scene is a common scene difficult to reasonably process in automatic driving.
When the main vehicle changes the road in the congested scene, if the algorithm is not properly processed, collision or even dangerous accidents are easy to happen. If the automatic driving algorithm has a large amount of data of lane changing of the main vehicle in a congested scene, the automatic driving algorithm can be subjected to simulation test, problems can be found as early as possible in the algorithm development stage, and the quality of the automatic driving algorithm can be better evaluated.
In addition, after a large amount of data of lane changing of the main vehicle in a congested scene is possessed, the data can be used as samples, manual marking is carried out on the samples to obtain a large amount of training data, and then the data is adopted to train the neural network model of automatic driving, so that the automatic driving can be finally more intelligently adapted to the scenes.
Disclosure of Invention
In view of this, the present application provides a method and an apparatus for collecting lane change data of a host vehicle in a congested scene, so as to collect the lane change data of the host vehicle in the congested scene.
In order to achieve the above object, a first aspect of the present application provides a method for collecting data of a main lane change in a congested scene, including:
acquiring automatic driving data of a main vehicle, wherein the automatic driving data comprises data under various themes;
dividing the automatic driving data from a time dimension to obtain each data frame taking time as an index, wherein the data frame comprises data under each theme in the belonged time;
combining semantic map data, and screening out target data frames of the main vehicle with lane change in a congested scene from all data frames;
and aggregating the target data frames according to preset conditions to obtain lane change data in a congested scene.
Preferably, the step of screening out a target data frame of the lane change of the host vehicle in the congested scene from the data frames by combining the semantic map data includes:
acquiring a neighboring occupation ratio of the data frames according to the semantic map data and the obstacle information and the host vehicle information in each data frame, wherein the neighboring occupation ratio is used for representing the occupation condition of a host vehicle neighboring area in the corresponding data frame;
screening out data frames under a congestion scene from all the data frames based on the adjacent occupancy ratio to obtain a candidate data frame set;
and acquiring a target data frame of the main vehicle with a lane change according to the semantic map data and the main vehicle information in each data frame in the candidate data frame set.
Preferably, before obtaining the adjacent occupancy ratio of each data frame according to the semantic map data and the obstacle information and the host vehicle information in the data frame, the method further comprises:
determining data frames under the non-congestion scene according to the semantic map data and the obstacle information and the main vehicle information in each data frame;
and eliminating the data frames under the non-congestion scene from each data frame.
Preferably, the process of determining the data frames in the non-congested scene according to the semantic map data and the obstacle information and the host vehicle information in each data frame comprises:
according to the semantic map data and the host vehicle information in each data frame, acquiring the speed of a host vehicle in the data frame and an adjacent area in a preset range of the host vehicle;
acquiring the number of obstacles in the adjacent area in each data frame according to the obstacle information in the data frame;
and determining the data frames of which the speed of the main vehicle is greater than a preset speed threshold value and the number of obstacles is less than a preset number threshold value as the data frames in the non-congestion scene.
Preferably, the process of obtaining the proximity occupancy ratio of each data frame according to the semantic map data and the obstacle information and the host vehicle information in the data frame includes:
for each data frame:
acquiring a drivable area and a surrounding area of the main vehicle according to semantic map data;
acquiring a first occupied area of a main vehicle outline according to the main vehicle information in the data frame;
acquiring a second occupied area of the obstacle outline according to the obstacle information in the data frame;
and acquiring the adjacent occupancy ratio of the data frame according to the travelable area, the surrounding area of the main vehicle, the first occupancy area and the second occupancy area.
Preferably, the process of acquiring the proximity occupancy ratio of the data frame based on the travelable region, the surrounding region of the host vehicle, the first occupancy region, and the second occupancy region includes:
calculating the intersection of the drivable area and the area around the main vehicle to obtain an adjacent drivable area;
combining the first occupied area and the second occupied area to obtain a total occupied area;
calculating the intersection of the adjacent travelable area and the total occupied area to obtain the adjacent occupied area;
and determining the ratio of the adjacent occupied area to the adjacent travelable area as the adjacent occupancy ratio.
Preferably, the process of acquiring a target data frame of a lane change in a host vehicle according to semantic map data and host vehicle information in each data frame in a candidate data frame set comprises:
aiming at each data frame in a candidate data frame set, acquiring a lane where a host vehicle is located in the data frame according to a semantic map and host vehicle information in the data frame;
selecting data frames from a candidate data frame set by using a preset sliding window and a preset sliding step length to obtain a plurality of sliding window objects, wherein the sliding window objects comprise a plurality of data frames;
and determining a sliding window object with a lane change according to the lane where the host vehicle is located in each data frame in the sliding window object, and determining the data frame in the sliding window object with the lane change as the target data frame.
Preferably, the process of acquiring the automated driving data of the host vehicle includes:
obtaining a bag file recorded by the main vehicle in the driving process, wherein the bag file comprises automatic driving data.
Preferably, the process of dividing the automatic driving data from a time dimension to obtain each data frame with time as an index includes:
dividing the time axis of the automatic driving data into a plurality of continuous time intervals by preset time steps;
aiming at data under each theme of automatic driving data, dividing the data under the theme into a plurality of time blocks under the theme by taking each time interval as a boundary according to the recording time of the data;
for each topic, if the time blocks under the topic are discontinuous in time, generating the time blocks of the time intervals according to the data recorded by a preset number of time blocks before and/or after the time intervals aiming at the time intervals without data records;
the time blocks of each topic under the same time interval form a data frame under the time interval.
Preferably, the process of aggregating the target data frames according to a preset condition to obtain data of lane change in a congested scene includes:
for each target data frame, aggregating the target data frames with the interval within a preset time threshold into a first candidate data packet;
acquiring the time span of each first candidate data packet, and determining the first candidate data packet with the time span smaller than a preset threshold value as invalid data;
and eliminating the invalid data from each first candidate data packet to obtain the data of lane change in the congested scene.
The second aspect of the present application provides a device for collecting data of a main lane change in a congested scene, including:
a data collection unit for acquiring automated driving data of a host vehicle, the automated driving data including data under various themes;
the data preprocessing unit is used for dividing the automatic driving data from a time dimension to obtain each data frame taking time as an index, and the data frame comprises data under each theme in the belonged time;
the data screening unit is used for screening out a target data frame of the main vehicle lane change under the congested scene from all the data frames by combining semantic map data;
and the data aggregation unit is used for aggregating the target data frames according to preset conditions to obtain the data of lane change in the congested scene.
The third aspect of the present application provides a device for collecting data of a main lane change in a congested scene, including: a memory and a processor;
the memory is used for storing programs;
the processor is used for executing the program to realize the steps of the main lane change data collection method in the congestion scene.
A third aspect of the present application provides a storage medium having stored thereon a computer program comprising: when being executed by a processor, the computer program realizes the steps of the main lane change data collection method in the congestion scene.
According to the technical scheme, a large amount of original driving data can be obtained by acquiring the automatic driving data of the main vehicle. For the automatic driving data of the main vehicle, the automatic driving data are divided from the time dimension to obtain each data frame taking time as an index. Wherein each data frame includes the autopilot data for the associated time. By the above-described division operation, frame alignment of the automated driving data is realized, so that the automated driving data in the specified time interval can be acquired through the time index, that is, various data of the host vehicle in the running process in the specified time can be aggregated, and the situation occurring in the running process in the specified time interval can be known conveniently.
And after obtaining each data frame, screening out a target data frame of the main vehicle lane change under the congested scene from each data frame by combining semantic map data. The target data frame may be a data frame that is not completely continuous or may be a data frame that is too short in duration. Therefore, the target data frames are aggregated according to preset conditions to obtain data of lane changing in the congested scene, and the data of lane changing in the congested scene relatively completely contains related information of each lane changing of the main vehicle in the congested scene.
By the lane changing data collection method in the congested scene, lane changing data of the main vehicle in the congested scene can be collected, and relevant data support is provided for development of an automatic driving technology.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below, it is obvious that the drawings in the following description are only embodiments of the present application, and for those skilled in the art, other drawings can be obtained according to the provided drawings without creative efforts.
FIG. 1 is a schematic diagram illustrating a method for collecting primary lane change data in a congested scene according to an embodiment of the present disclosure;
FIG. 2 illustrates a schematic diagram of a proximity occupancy of a host vehicle as disclosed in an embodiment of the present application;
FIG. 3 is a schematic diagram of an apparatus of a method for collecting primary lane change data in a congested scene according to an embodiment of the present disclosure;
fig. 4 is a schematic diagram of a device for collecting main lane change data in a congested scene, which is disclosed in an embodiment of the present application.
Detailed Description
The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
The method for collecting main lane change data in a congested scene provided by the embodiment of the present application is described below. Referring to fig. 1, a method for collecting primary lane change data in a congested scene according to an embodiment of the present application may include the following steps:
step S100, automatic driving data of the main vehicle is obtained.
The host vehicle may generate various data during travel, which may include data of the host vehicle itself and data of obstacles around the host vehicle during travel. The data of the host vehicle itself is, for example, the position, speed, and the like of the host vehicle, and the data of surrounding obstacles is, for example, the number, position, speed, contour, and the like of the obstacles.
Alternatively, the data of the host vehicle itself and surrounding obstacles during driving can be obtained through sensors such as a GPS, an IMU, a camera and a laser radar, and then the information such as the type, the position, the speed and the like of the obstacles can be identified according to the data through artificial intelligence and other technologies. And finally, the recognized results are recorded in the vehicle-mounted mobile hard disk, so that the vehicle-mounted mobile hard disk is convenient to collect and process subsequently.
And step S200, carrying out structuralization processing on the automatic driving data to obtain each data frame.
Specifically, data under each topic of the automatic driving data are divided from a time dimension to obtain each data frame with time as an index, and the data frame comprises data under each topic in the belonged time.
The data of different subjects may have different timestamps, and the data under each subject is divided into data frames according to a certain time step, for example, 0.1 second, so that the data under each subject can be frame-aligned according to time. Further, referring to table 1, the divided data under each theme may also be stored in a table structure, and each line of data constitutes a data frame, which may be regarded as information generated at the same time.
Table 1: data frame structure
| Time | Subject 1: information of the host vehicle | Subject 2: obstacle information |
| 0.1s | 0.1-0.2 of the master vehicle information | Obstacle information between 0.1 and 0.2 |
| 0.2s | 0.2-0.3 of the master vehicle information | Obstacle information between 0.2 and 0.3 |
| 0.3s | 0.3-0.4 of the master vehicle information | 0.3-0.4 obstacleInformation |
And step S300, screening out target data frames of the main vehicle lane change under the congested scene from the data frames.
Specifically, a target data frame of the main vehicle lane changing under the congested scene is screened from all data frames by combining semantic map data.
The semantic map records road surface information at each geographic position, such as lane data at each geographic position and data information such as a nearby formalized area. Through the semantic map data, the host vehicle information and the obstacle information in each data frame, a scene with congestion can be inferred, and a scene with a lane change of the host vehicle can be determined. And integrating various inference results, and finally determining a target data frame of the main vehicle with the lane change in the congested scene from all the data frames.
And step S400, aggregating the target data frames according to preset conditions to obtain lane change data in a congested scene.
Specifically, the target data frame obtained in step S300 may be a data frame that is not completely continuous in time, or may be a fragmented data frame that is too short in duration. The target data frames also need to be aggregated to obtain more reasonable data.
The data of lane change in the congested scene obtained through aggregation relatively completely contains the relevant information of each lane change of the main vehicle in the congested scene.
According to the embodiment of the application, a large amount of original driving data can be obtained by acquiring the automatic driving data of the main vehicle. For the automatic driving data of the main vehicle, the automatic driving data are divided from the time dimension to obtain each data frame taking time as an index. Wherein each data frame includes the autopilot data for the associated time. By the above-described division operation, frame alignment of the automated driving data is realized, so that the automated driving data in the specified time interval can be acquired through the time index, that is, various data of the host vehicle in the running process in the specified time can be aggregated, and the situation occurring in the running process in the specified time interval can be known conveniently.
And after each data frame is obtained, screening out a target data frame of the main vehicle lane change under the congested scene from each data frame by combining semantic map data. The target data frame may be a data frame that is not completely continuous or may be a data frame that is too short in duration. Therefore, the target data frame is aggregated according to the preset conditions to obtain the lane change data in the congested scene, and the lane change data in the congested scene relatively completely contains the relevant information of each lane change of the main vehicle in the congested scene.
By the lane changing data collection method in the congested scene, lane changing data of the main vehicle in the congested scene can be collected, and relevant data support is provided for development of an automatic driving technology.
In some embodiments of the present application, the process of acquiring the automatic driving data of the host vehicle in step S100 may include:
obtaining a bag file recorded by the main vehicle in the driving process, wherein the bag file comprises automatic driving data.
Specifically, the host vehicle can generate various data during traveling, and the data can be saved in the form of bag files in the vehicle-mounted mobile hard disk. When the running data of the host vehicle needs to be acquired, the corresponding running data can be obtained from a bag file recorded by the host vehicle during running.
The bag file may be a bag file in a Robot Operating System (ROS), or a file in another format used in the unmanned technology.
In some embodiments of the present application, the step S200 of dividing the data under each topic of the automatic driving data from a time dimension to obtain each data frame with time as an index may include:
s1, dividing the time axis of the data under each topic into a plurality of continuous time intervals with a preset time step.
And S2, dividing the data under the theme into a plurality of time blocks under the theme by taking each time interval as a boundary according to the recording time of the data for each theme.
And S3, for each theme, if the time blocks under the theme are not continuous in time, for the time interval without data record, generating the time blocks of the time interval according to the data recorded by the preset number of time blocks before and/or after the time interval.
In particular, there may be relatively large differences in the acquisition periods of data under different subjects. For example, for a host vehicle velocity, the period of data acquisition may be quasi-real-time, so its data frames are substantially continuous; the data acquisition of the obstacle around the host vehicle may be performed at intervals (e.g., 5 seconds). Therefore, there may be a problem that the data frames in some subjects are discontinuous in time. At this time, the missing data frame may be subjected to frame complementing by the method in a 3. In general, missing data frames may be frame-complemented by the content of data frames preceding or following the missing time period.
And the time blocks of all subjects in the same time interval form a data frame in the time interval.
In some embodiments of the present application, for each data frame obtained after the processing in step S200, the data frame may also be stored in a column storage manner in a distributed storage of a cloud or a data center, so that the data may be processed in parallel by using a distributed computing system. Specifically, the calculation tasks may be distributed to multiple computers, the multiple calculation tasks are executed simultaneously, and each calculation task acquires a batch of data frames to perform the data processing in step S300, so as to increase the operation speed.
In some embodiments of the present application, the step S300 of screening out, from the respective data frames, a target data frame of the host vehicle that has a lane change in a congested scene in combination with the semantic map data may include:
s1, according to the semantic map data and the obstacle information and the host vehicle information in each data frame, acquiring the adjacent occupancy ratio of the data frame, wherein the adjacent occupancy ratio can be used for representing the occupancy condition of the host vehicle adjacent area in the corresponding data frame.
For example, through semantic map data, the ground condition information of the position of the host vehicle in a certain data frame can be obtained; the ground condition information is combined with the obstacle information and the main vehicle information in the data frame, so that the occupation condition of the adjacent area of the position of the main vehicle can be calculated.
And S2, screening the data frames under the congestion scene from the data frames based on the adjacent occupancy ratio to obtain a candidate data frame set.
The candidate data frame set comprises data frames in a congestion scene.
S3, according to the semantic map data and the host vehicle information in each data frame in the candidate data frame set, obtaining the target data frame of the host vehicle with the lane change.
Specifically, the semantic map data records road surface information at each geographic location, such as lane data information at each geographic location. The position of the host vehicle can be acquired by host vehicle information in a certain data frame. Combining the position and the semantic map data, the lane where the host vehicle is located in the data frame can be obtained. For several consecutive data frames, if the lane of the host vehicle changes, the host vehicle may be considered to have performed a lane change operation.
In some embodiments of the present application, before the step S1 obtaining the adjacent occupancy ratio of each data frame according to the semantic map data and the obstacle information and the host vehicle information in the data frame, the method may further include:
and S5, determining the data frames in the non-congestion scene according to the semantic map data and the obstacle information and the main vehicle information in each data frame.
And S6, removing the data frames in the non-congestion scene from the data frames.
For example, for a certain data frame, the vicinity of the host vehicle may be determined from the semantic map data and the position information in the host vehicle information. In combination with the position information in the obstacle information, an obstacle in the vicinity of the host vehicle can be determined. Further, the congestion condition of the area near the host vehicle can be determined based on information such as the number of obstacles in the area near the host vehicle, the size of each obstacle, and the like.
For another example, for a certain data frame, it is possible to know whether the host vehicle is in a fast-traveling state or a slow-creeping or stopped state from the host vehicle speed in the host vehicle information. Further, based on these traveling states of the host vehicle, it is possible to deduce whether the road is congested. Generally, when traffic congestion occurs, the speed of the host vehicle is relatively greatly restricted; when the host vehicle can travel at high speed, the road surface condition is considered to be uncongested.
The determination of the congestion condition may have relatively large errors, so that the data frame of the congestion scene cannot be determined simply by counting the obstacle conditions and the vehicle speed of the nearby area, but a large amount of data frames of the non-congestion scene can be filtered by the method.
Based on this, in some embodiments of the present application, the step of determining the data frame in the non-congested scene according to the semantic map data and the obstacle information and the host vehicle information in each data frame at S5 may include:
s51, acquiring the speed of the host in the data frame and the adjacent area in the preset range of the host according to the semantic map data and the host information in each data frame.
S52, acquiring the number of obstacles in the adjacent area in the data frame according to the obstacle information in each data frame.
And S53, determining the data frames with the speed of the host vehicle larger than a preset speed threshold value and the number of obstacles smaller than a preset number threshold value as the data frames in the non-congestion scene.
By the method of the embodiment, a large number of data frames of the non-congestion scene can be filtered, so that the subsequent calculation amount is reduced, and the calculation efficiency is effectively improved.
In some embodiments of the present application, the step of obtaining the proximity occupancy ratio of each data frame according to the semantic map data and the obstacle information and the host vehicle information in the data frame at S1 may include:
for each data frame:
and S11, acquiring a drivable area and a surrounding area of the main vehicle according to the semantic map data.
Specifically, the semantic map recording data includes a travelable area near each geographic position, and the travelable area near the position point of the host vehicle is queried in the semantic map data by the position information of the host vehicle.
The area around the host vehicle may also be determined based on the position of the host vehicle. For example, referring to fig. 2, a circular area with a radius R is taken as a surrounding area of the main car with the ground as a plane and the main car as a center. Wherein, R is a preset numerical value.
S12, according to the host vehicle information in the data frame, a first occupation area of the host vehicle outline is obtained.
And S13, acquiring a second occupied area of the obstacle outline according to the obstacle information in the data frame.
S14, acquiring a proximity occupancy ratio of the data frame based on the travelable region, the surrounding area of the host vehicle, the first occupancy region, and the second occupancy region.
The step is that the adjacent occupancy ratio is obtained by obtaining the drivable area of the main vehicle, the obstacles in the drivable area around the main vehicle and the area occupied by the main vehicle in a certain data frame, and the adjacent occupancy ratio reflects the occupation condition of the adjacent area of the main vehicle in the corresponding data frame.
In some embodiments of the present application, the step of obtaining the proximity occupancy ratio of the data frame according to the travelable region, the surrounding area of the host vehicle, the first occupancy region, and the second occupancy region at S14 may include:
s141, calculating the intersection of the travelable area and the surrounding area of the main vehicle to obtain the adjacent travelable area.
And S142, combining the first occupied area and the second occupied area to obtain a total occupied area.
And S143, calculating the intersection of the adjacent travelable area and the total occupied area to obtain the adjacent occupied area.
S144, determining a ratio of the adjacent occupied area to the adjacent travelable area as the adjacent occupancy ratio.
For example, referring to fig. 2, assuming that the travelable region is D1, the surrounding area of the host vehicle is D2, and assuming that the first occupied region is D3 and the second occupied region is D4, the adjacent travelable region D5 can be represented as:
D5＝D1∩D2
the total occupied area D6 may be expressed as:
D6＝D3∪D4
the proximity occupancy area D7 may be expressed as:
D7＝D5∩D6
assuming that the area of the adjacent travelable region D5 is M and the area of the adjacent occupied region D7 is N, the adjacent occupancy ratio P can be expressed as:
P＝N/M
in some embodiments of the present application, the step S3 of obtaining the target data frame of the lane change of the host vehicle according to the semantic map data and the host vehicle information in each data frame of the candidate data frame set may include:
s31, aiming at each data frame in the candidate data frame set, acquiring a lane where the host vehicle is located in the data frame according to the semantic map and the host vehicle information in the data frame.
And S32, selecting data frames from the candidate data frame set according to a preset sliding window and a preset sliding step length to obtain a plurality of sliding window objects, wherein the sliding window objects comprise a plurality of data frames.
S33, according to the lane where the main vehicle is located in each data frame in the sliding window object, determining the sliding window object with the lane change, and determining the data frame in the sliding window object with the lane change as the target data frame.
For example, according to the above-described S31, the lane in which the host vehicle is located in each data frame in the candidate data frame set is acquired. And setting a sliding window with the length of 4 seconds, and selecting the data frame from the candidate data frame set by a sliding step of 1 second. The foregoing 4 seconds and 1 second are examples, and different sliding window lengths and sliding step lengths may be set empirically in practical applications. Assuming that the candidate data frame is a continuous data frame from time t to t being 50s to 100s, the obtained sliding window object includes:
(50，54)、(51，55)、…、(95，99)，(96，100)
the sliding window object (50, 54) includes candidate frames in a time interval from time t to 54s, the sliding window object (51, 55) includes candidate frames in a time interval from time t to 55s, and so on.
Assuming that the lane where the host vehicle is located in the data frame at the time t-60 is x, and the lane where the host vehicle is located in the data frame at the time t-61 is y, where x and y are different lanes, it may be determined that the following sliding window object has caused a lane change of the host vehicle:
(57，61)、(58，62)、(59，63)、(60，64)
then, the subsequent data frame in the time interval from t 57 to t 64 may be determined as the target data frame.
In some embodiments of the present application, the preset condition mentioned in the above step S400 may include at least one of:
1) aggregating target data frames with intervals within a preset time threshold value into a channel-changing data under a congestion scene;
2) acquiring a missing data frame for a channel change data under a congestion scene with frame missing, and aggregating the missing data frame into the channel change data under the congestion scene;
3) and confirming the data of lane change under the congestion scene with the total time span smaller than the preset threshold as invalid data.
Based on this, in some embodiments of the application, the process of aggregating the target data frames according to the preset condition in step S400 to obtain the lane change data in the congested scene may include:
and S1, for each target data frame, aggregating the target data frames with the interval within the preset time threshold into a first candidate data packet.
And S2, acquiring the time span of each first candidate data packet, and determining the first candidate data packet with the time span smaller than a preset threshold value as invalid data.
And S3, removing the invalid data from each first candidate data packet to obtain data for lane change in a congested scene.
In S1, there may be a case where the data frame is discontinuous for a certain first candidate packet. Based on this, in some embodiments of the present application, the above S1 may include:
for the first candidate data packet with discontinuous data frames, according to the time interval with data frame missing, the data frames in the time interval can be extracted from each data frame, and the data frames are added to the first candidate data packet.
The following describes the device for collecting lane change data of a host in a congested scene, and the device for collecting lane change data of a host in a congested scene described below and the method for collecting lane change data of a host in a congested scene described above may be referred to in correspondence.
Referring to fig. 3, an apparatus for collecting data of a main lane change in a congested scene according to an embodiment of the present disclosure may include:
a data collection unit 21 for acquiring automated driving data of the host vehicle, the automated driving data including data under respective subjects;
the data preprocessing unit 22 is configured to divide the automatic driving data from a time dimension to obtain data frames indexed by time, where the data frames include data under each topic in the time to which the data frames belong;
the data screening unit 23 is used for screening out a target data frame of the main vehicle lane change in a congested scene from each data frame by combining semantic map data;
and the data aggregation unit 24 is configured to aggregate the target data frames according to preset conditions to obtain data of lane change in a congested scene.
The process of the data screening unit 23, in combination with the semantic map data, screening out a target data frame of the host lane change in the congested scene from the data frames may include:
acquiring a neighboring occupation ratio of the data frames according to the semantic map data and the obstacle information and the host vehicle information in each data frame, wherein the neighboring occupation ratio is used for representing the occupation condition of a host vehicle neighboring area in the corresponding data frame;
screening out data frames under a congestion scene from all the data frames based on the adjacent occupancy ratio to obtain a candidate data frame set;
and acquiring a target data frame of the main vehicle with a lane change according to the semantic map data and the main vehicle information in each data frame in the candidate data frame set.
The data filtering unit 23 may further include, before acquiring the adjacent occupancy ratio of each data frame according to the semantic map data and the obstacle information and the host vehicle information in the data frame:
determining data frames under the non-congestion scene according to the semantic map data and the obstacle information and the main vehicle information in each data frame;
and eliminating the data frames under the non-congestion scene from each data frame.
The process of determining the data frame in the non-congestion scene by the data screening unit 23 according to the semantic map data and the obstacle information and the host vehicle information in each data frame may include:
according to the semantic map data and the host vehicle information in each data frame, acquiring the speed of a host vehicle in the data frame and an adjacent area in a preset range of the host vehicle;
acquiring the number of obstacles in the adjacent area in each data frame according to the obstacle information in the data frame;
and determining the data frames of which the speed of the main vehicle is greater than a preset speed threshold value and the number of obstacles is less than a preset number threshold value as the data frames in the non-congestion scene.
The process of the data screening unit 23 obtaining the adjacent occupancy ratio of each data frame according to the semantic map data and the obstacle information and the host vehicle information in the data frame may include:
for each data frame:
acquiring a drivable area and a surrounding area of the main vehicle according to semantic map data;
acquiring a first occupied area of a main vehicle outline according to the main vehicle information in the data frame;
acquiring a second occupied area of the obstacle outline according to the obstacle information in the data frame;
and acquiring the adjacent occupancy ratio of the data frame according to the travelable area, the surrounding area of the main vehicle, the first occupancy area and the second occupancy area.
The process of the data filtering unit 23 acquiring the proximity occupancy ratio of the data frame according to the travelable region, the surrounding region of the host vehicle, the first occupancy region, and the second occupancy region may include:
calculating the intersection of the drivable area and the area around the main vehicle to obtain an adjacent drivable area;
combining the first occupied area and the second occupied area to obtain a total occupied area;
calculating the intersection of the adjacent travelable area and the total occupied area to obtain the adjacent occupied area;
and determining the ratio of the adjacent occupied area to the adjacent travelable area as the adjacent occupancy ratio.
The process of obtaining the target data frame of which the host lane is changed by the data screening unit 23 according to the semantic map data and the host information in each data frame in the candidate data frame set may include:
aiming at each data frame in a candidate data frame set, acquiring a lane where a host vehicle is located in the data frame according to a semantic map and host vehicle information in the data frame;
selecting data frames from a candidate data frame set by using a preset sliding window and a preset sliding step length to obtain a plurality of sliding window objects, wherein the sliding window objects comprise a plurality of data frames;
and determining a sliding window object with a lane change according to the lane where the host vehicle is located in each data frame in the sliding window object, and determining the data frame in the sliding window object with the lane change as the target data frame.
The process of the data collection unit 21 acquiring the automated driving data of the host vehicle may include:
obtaining a bag file recorded by the main vehicle in the driving process, wherein the bag file comprises automatic driving data.
The process of the data preprocessing unit 22 dividing the automatic driving data from the time dimension to obtain each data frame with time as an index may include:
dividing a time axis of data under each theme of the automatic driving data into a plurality of continuous time intervals by preset time steps;
for data under each theme, dividing the data under the theme into a plurality of time blocks under the theme by taking each time interval as a boundary according to the recording time of the data;
for each topic, if the time blocks under the topic are discontinuous in time, generating the time blocks of the time intervals according to the data recorded by a preset number of time blocks before and/or after the time intervals aiming at the time intervals without data records;
the time blocks of each topic under the same time interval form a data frame under the time interval.
The process of aggregating the target data frame by the data aggregation unit 24 according to preset conditions to obtain lane change data in a congested scene includes:
for each target data frame, aggregating the target data frames with the interval within a preset time threshold into a first candidate data packet;
acquiring the time span of each first candidate data packet, and determining the first candidate data packets with the time span smaller than a preset threshold value as invalid data;
and eliminating the invalid data from each first candidate data packet to obtain the data of lane change in the congested scene.
The device for collecting the main lane change data in the congested scene can be applied to equipment for collecting the main lane change data in the congested scene. Alternatively, fig. 4 is a block diagram illustrating a hardware structure of a device for collecting primary lane change data in a congested scene, and referring to fig. 4, the hardware structure of the device for collecting primary lane change data in a congested scene may include: at least one processor 31, at least one communication interface 32, at least one memory 33 and at least one communication bus 34.
In the embodiment of the present application, the number of the processor 31, the communication interface 32, the memory 33 and the communication bus 34 is at least one, and the processor 31, the communication interface 32 and the memory 33 complete communication with each other through the communication bus 34;
the processor 31 may be a central processing unit CPU, or an application Specific Integrated circuit asic, or one or more Integrated circuits configured to implement embodiments of the present application, etc.;
the memory 32 may comprise a high-speed RAM memory, and may further comprise a non-volatile memory (non-volatile memory) or the like, such as at least one disk memory;
wherein the memory 33 stores a program and the processor 31 may invoke the program stored in the memory 33, the program being for:
acquiring automatic driving data of a main vehicle, wherein the automatic driving data comprises data under various themes;
dividing the automatic driving data from a time dimension to obtain each data frame taking time as an index, wherein the data frame comprises data under each theme in the belonged time;
combining semantic map data, and screening out target data frames of the main vehicle with lane change in a congested scene from all data frames;
and aggregating the target data frames according to preset conditions to obtain lane change data in a congested scene.
Alternatively, the detailed function and the extended function of the program may be as described above.
Embodiments of the present application further provide a storage medium, where a program suitable for execution by a processor may be stored, where the program is configured to:
acquiring automatic driving data of a main vehicle;
dividing the automatic driving data from a time dimension to obtain each data frame taking time as an index, wherein the data frame comprises data under each theme in the belonged time;
combining semantic map data, and screening out target data frames of the main vehicle with lane change in a congested scene from all data frames;
and aggregating the target data frames according to preset conditions to obtain lane change data in a congested scene.
Alternatively, the detailed function and the extended function of the program may be as described above.
In summary, the following steps:
according to the method and the device, a large amount of original driving data can be obtained by acquiring the automatic driving data of the main vehicle. For the automatic driving data of the main vehicle, the automatic driving data are divided from the time dimension to obtain each data frame taking time as an index. Wherein each data frame includes the autopilot data for the associated time. By the above-described division operation, frame alignment of the automated driving data is realized, so that the automated driving data in the specified time interval can be acquired through the time index, that is, various data of the host vehicle in the running process in the specified time can be aggregated, and the situation occurring in the running process in the specified time interval can be known conveniently.
And after obtaining each data frame, screening out a target data frame of the main vehicle lane change under the congested scene from each data frame by combining semantic map data. The target data frame may be a data frame that is not completely continuous or may be a data frame that is too short in duration. Therefore, the target data frames are aggregated according to preset conditions to obtain data of lane changing in the congested scene, and the data of lane changing in the congested scene relatively completely contains related information of each lane changing of the main vehicle in the congested scene.
By the lane changing data collection method in the congested scene, lane changing data of the main vehicle in the congested scene can be collected, and relevant data support is provided for development of an automatic driving technology.
Finally, it should also be noted that, in this document, relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
The embodiments in the present description are described in a progressive manner, each embodiment focuses on differences from other embodiments, the embodiments may be combined as needed, and the same and similar parts may be referred to each other.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (12)
1. A method for collecting main lane change data in a congested scene is characterized by comprising the following steps:
acquiring automatic driving data of a main vehicle, wherein the automatic driving data comprises data under various themes;
dividing the automatic driving data from a time dimension to obtain each data frame taking time as an index, wherein the data frame comprises data under each theme in the belonged time;
combining semantic map data, and screening out target data frames of the main vehicle with lane change in a congested scene from all data frames;
aggregating the target data frames according to preset conditions to obtain lane change data in a congested scene;
the process of screening out the target data frame of the main vehicle lane change under the congested scene from all the data frames by combining the semantic map data comprises the following steps:
acquiring a neighboring occupation ratio of the data frames according to the semantic map data and the obstacle information and the host vehicle information in each data frame, wherein the neighboring occupation ratio is used for representing the occupation condition of a host vehicle neighboring area in the corresponding data frame;
screening out data frames under a congestion scene from all the data frames based on the adjacent occupancy ratio to obtain a candidate data frame set;
and acquiring a target data frame of the main vehicle with a lane change according to the semantic map data and the main vehicle information in each data frame in the candidate data frame set.
2. The method of claim 1, further comprising, prior to obtaining the proximity occupancy ratio for each data frame based on semantic map data and obstacle information and host vehicle information in the data frame:
determining data frames under the non-congestion scene according to the semantic map data and the obstacle information and the main vehicle information in each data frame;
and eliminating the data frames under the non-congestion scene from each data frame.
3. The method of claim 2, wherein determining data frames in non-congested scenes from semantic map data and obstacle information and host vehicle information in each data frame comprises:
according to the semantic map data and the host vehicle information in each data frame, acquiring the speed of a host vehicle in the data frame and an adjacent area in a preset range of the host vehicle;
acquiring the number of obstacles in the adjacent area in each data frame according to the obstacle information in the data frame;
and determining the data frames of which the speed of the main vehicle is greater than a preset speed threshold value and the number of obstacles is less than a preset number threshold value as the data frames in the non-congestion scene.
4. The method according to claim 1, wherein the process of obtaining the proximity occupancy ratio of the data frames from the semantic map data and the obstacle information and the host vehicle information in each data frame comprises:
for each data frame:
acquiring a drivable area and a surrounding area of the main vehicle according to semantic map data;
acquiring a first occupied area of a main vehicle outline according to the main vehicle information in the data frame;
acquiring a second occupied area of the obstacle outline according to the obstacle information in the data frame;
and acquiring the adjacent occupancy ratio of the data frame according to the travelable area, the surrounding area of the main vehicle, the first occupancy area and the second occupancy area.
5. The method according to claim 4, wherein the process of acquiring the proximity occupancy ratio of the data frame based on the travelable region, the surrounding area of the host vehicle, the first occupancy area, and the second occupancy area includes:
calculating the intersection of the travelable area and the surrounding area of the main vehicle to obtain an adjacent travelable area;
combining the first occupied area and the second occupied area to obtain a total occupied area;
calculating the intersection of the adjacent travelable area and the total occupied area to obtain an adjacent occupied area;
and determining the ratio of the adjacent occupied area to the adjacent travelable area as the adjacent occupancy ratio.
6. The method of claim 1, wherein obtaining a target data frame for which a host lane change has occurred based on the semantic map data and host information in each data frame of the set of candidate data frames comprises:
aiming at each data frame in a candidate data frame set, acquiring a lane where a host vehicle is located in the data frame according to a semantic map and host vehicle information in the data frame;
selecting data frames from a candidate data frame set by using a preset sliding window and a preset sliding step length to obtain a plurality of sliding window objects, wherein the sliding window objects comprise a plurality of data frames;
and determining a sliding window object with a lane change according to the lane where the host vehicle is located in each data frame in the sliding window object, and determining the data frame in the sliding window object with the lane change as the target data frame.
7. The method of claim 1, wherein the process of obtaining automated driving data for the host vehicle comprises:
obtaining a bag file recorded by the main vehicle in the driving process, wherein the bag file comprises automatic driving data.
8. The method of claim 1, wherein the act of partitioning the autopilot data from a time dimension into individual frames of data indexed by time comprises:
dividing the time axis of the automatic driving data into a plurality of continuous time intervals by preset time steps;
for data under each theme of the automatic driving data, dividing the data under the theme into a plurality of time blocks under the theme by taking each time interval as a boundary according to the recording time of the data;
for each topic, if the time blocks under the topic are discontinuous in time, generating the time blocks of the time intervals according to the data recorded by a preset number of time blocks before and/or after the time intervals aiming at the time intervals without data records;
the time blocks of each topic in the same time interval constitute a data frame in the time interval.
9. The method as claimed in claim 1, wherein the process of aggregating the target data frames according to a preset condition to obtain data of lane change in a congested scene includes:
for each target data frame, aggregating the target data frames with the interval within a preset time threshold into a candidate data packet;
and acquiring the time span of each candidate data packet, and determining the candidate data packets with the time span larger than a preset threshold value as the data of lane change in the congested scene.
10. A device for collecting main lane change data in a congested scene is characterized by comprising:
a data collection unit for acquiring automated driving data of a host vehicle, the automated driving data including data under various themes;
the data preprocessing unit is used for dividing the automatic driving data from a time dimension to obtain each data frame taking time as an index, and the data frame comprises data under each theme in the belonged time;
the data screening unit is used for screening out a target data frame of the main vehicle lane change under the congested scene from all the data frames by combining semantic map data;
the data aggregation unit is used for aggregating the target data frame according to preset conditions to obtain lane change data in a congested scene;
the data screening unit is combined with semantic map data to screen out target data frames of the main vehicle lane changing under the congested scene from all the data frames, and the process comprises the following steps:
acquiring a neighboring occupation ratio of the data frames according to the semantic map data and the obstacle information and the host vehicle information in each data frame, wherein the neighboring occupation ratio is used for representing the occupation condition of a host vehicle neighboring area in the corresponding data frame;
screening out data frames under a congestion scene from all the data frames based on the adjacent occupancy ratio to obtain a candidate data frame set;
and acquiring a target data frame of the main vehicle with a lane change according to the semantic map data and the main vehicle information in each data frame in the candidate data frame set.
11. The utility model provides a collection equipment of main car lane change data under scene of blocking up which characterized in that includes: a memory and a processor;
the memory is used for storing programs;
the processor is used for executing the program to realize the steps of the main lane change data collection method in the congestion scene according to any one of claims 1 to 9.
12. A storage medium having stored thereon a computer program, wherein the computer program, when executed by a processor, performs the steps of the method for collecting primary lane change data in a congested scene as recited in any of claims 1-9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111007467.5A CN113706870B (en) | 2021-08-30 | 2021-08-30 | Method for collecting main vehicle lane change data in congested scene and related equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111007467.5A CN113706870B (en) | 2021-08-30 | 2021-08-30 | Method for collecting main vehicle lane change data in congested scene and related equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113706870A CN113706870A (en) | 2021-11-26 |
| CN113706870B true CN113706870B (en) | 2022-06-10 |
Family
ID=78657099
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111007467.5A Active CN113706870B (en) | 2021-08-30 | 2021-08-30 | Method for collecting main vehicle lane change data in congested scene and related equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113706870B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114547403B (en) * | 2021-12-30 | 2023-05-23 | 广州文远知行科技有限公司 | Method, device, equipment and storage medium for collecting variable-track scene |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20170140024A (en) * | 2016-06-12 | 2017-12-20 | 이동은 | The system to reduce traffic congestion by using artificial intelligence to collect traffic information |
| CN107742418A (en) * | 2017-09-29 | 2018-02-27 | 东南大学 | A kind of urban expressway traffic congestion status and stifled point position automatic identifying method |
| CN110020797A (en) * | 2019-03-27 | 2019-07-16 | 清华大学苏州汽车研究院（吴江） | The evaluation method of automatic Pilot test scene based on perception defect |
| WO2020184084A1 (en) * | 2019-03-14 | 2020-09-17 | オムロン株式会社 | Learning device, deduction device, data generation device, learning method, and learning program |
Family Cites Families (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8996234B1 (en) * | 2011-10-11 | 2015-03-31 | Lytx, Inc. | Driver performance determination based on geolocation |
| JP6558239B2 (en) * | 2015-12-22 | 2019-08-14 | アイシン・エィ・ダブリュ株式会社 | Automatic driving support system, automatic driving support method, and computer program |
| US10983524B2 (en) * | 2018-04-12 | 2021-04-20 | Baidu Usa Llc | Sensor aggregation framework for autonomous driving vehicles |
| US10816984B2 (en) * | 2018-04-13 | 2020-10-27 | Baidu Usa Llc | Automatic data labelling for autonomous driving vehicles |
| CN113642633B (en) * | 2018-06-11 | 2023-06-20 | 百度在线网络技术（北京）有限公司 | Method, device, equipment and medium for classifying driving scene data |
| CN111325230B (en) * | 2018-12-17 | 2023-09-12 | 上海汽车集团股份有限公司 | Online learning method and online learning device for vehicle lane change decision model |
| CN111880533B (en) * | 2020-07-16 | 2023-03-24 | 华人运通(上海)自动驾驶科技有限公司 | Driving scene reconstruction method, device, system, vehicle, equipment and storage medium |
| CN111915915A (en) * | 2020-07-16 | 2020-11-10 | 华人运通(上海)自动驾驶科技有限公司 | Driving scene reconstruction method, device, system, vehicle, equipment and storage medium |
| CN111932884A (en) * | 2020-08-13 | 2020-11-13 | 中国第一汽车股份有限公司 | Traffic jam recognition method and device, vehicle and storage medium |
| CN111951583A (en) * | 2020-08-31 | 2020-11-17 | 维沃移动通信有限公司 | Prompting method and electronic equipment |
| CN112417953B (en) * | 2020-10-12 | 2022-07-19 | 腾讯科技（深圳）有限公司 | Road condition detection and map data updating method, device, system and equipment |
| CN112632304B (en) * | 2020-12-31 | 2024-05-03 | 广州文远知行科技有限公司 | Index-based data searching method, device, server and storage medium |
-
2021
- 2021-08-30 CN CN202111007467.5A patent/CN113706870B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20170140024A (en) * | 2016-06-12 | 2017-12-20 | 이동은 | The system to reduce traffic congestion by using artificial intelligence to collect traffic information |
| CN107742418A (en) * | 2017-09-29 | 2018-02-27 | 东南大学 | A kind of urban expressway traffic congestion status and stifled point position automatic identifying method |
| WO2020184084A1 (en) * | 2019-03-14 | 2020-09-17 | オムロン株式会社 | Learning device, deduction device, data generation device, learning method, and learning program |
| CN110020797A (en) * | 2019-03-27 | 2019-07-16 | 清华大学苏州汽车研究院（吴江） | The evaluation method of automatic Pilot test scene based on perception defect |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113706870A (en) | 2021-11-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20200211374A1 (en) | System, method, and apparatus for analyzing a traffic road condition | |
| CN110751828B (en) | Road congestion measuring method and device, computer equipment and storage medium | |
| Al Mamun et al. | An intelligent smartphone based approach using IoT for ensuring safe driving | |
| CN105355039A (en) | Road condition information processing method and equipment | |
| JP7051424B2 (en) | Anomaly detection device | |
| US11361555B2 (en) | Road environment monitoring device, road environment monitoring system, and road environment monitoring program | |
| CN106056910A (en) | Method and device for traffic state detection | |
| WO2024067210A1 (en) | Traffic state determination method and apparatus, and device | |
| CN113706870B (en) | Method for collecting main vehicle lane change data in congested scene and related equipment | |
| CN112700640B (en) | Road state monitoring method, server, vehicle-mounted equipment and road side equipment | |
| CN109661692B (en) | Traffic incident prediction method and device and terminal equipment | |
| EP4083959A1 (en) | Traffic flow machine-learning modeling system and method applied to vehicles | |
| CN112466118A (en) | Vehicle driving behavior recognition method, system, electronic device and storage medium | |
| CN116758762A (en) | Control method based on big data | |
| JP2020038595A (en) | Data collector, data collecting system, and method for collecting data | |
| CN114547403B (en) | Method, device, equipment and storage medium for collecting variable-track scene | |
| CN111949840B (en) | Method and device for constructing topological graph structure based on data of Internet of things | |
| CN114495505B (en) | Method, device, medium and server for predicting passing duration of congestion road section | |
| CN115257771A (en) | Intersection identification method, electronic device and storage medium | |
| CN114677836A (en) | Intersection multi-vehicle cooperation method and system based on minimum cluster coverage | |
| JP6835404B2 (en) | Probe information creation program, probe information creation method and probe information creation system | |
| CN112597924A (en) | Electric bicycle track tracking method, camera device and server | |
| JP6928082B2 (en) | Urban road running speed processing methods, urban road running speed processing equipment, equipment, and non-volatile computer storage media | |
| CN111599169A (en) | Road side unit inspection method, equipment and computer readable storage medium | |
| CN114582120B (en) | Hidden community discovery method, device, equipment and medium based on vehicle track |
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 |