Driving behavior analysis method, device and equipment of traffic monitoring system

Abstract

The invention provides a driving behavior analysis method, device and equipment for a traffic monitoring system, and the method comprises the steps: generating a track point of a traffic target based on the perception data of a multi-source sensor; assigning lane attributes to the track points; the traffic signal lamp state information is associated with the track points, and the traffic signal lamp state information is determined according to detection data of a traffic signal machine detection model and received traffic police signal machine data; when the number of the track points reaches a preset number, generating a motion track of the traffic target based on the track points; when the motion track comprises at least one inflection point, according to at least one preset multi-stage track identification strategy, driving behavior identification is carried out on the motion track, the driving behavior of the traffic target is determined, and the direction included angle of stage tracks on the two sides of the inflection point is larger than a first threshold value. The detection capability of a traffic monitoring system on illegal behaviors in a complex traffic scene is improved, and the efficiency and accuracy of traffic management are further improved.

Description

Technical Field

[0001] The present application relates to the technical field of traffic monitoring, and in particular to a driving behavior analysis method, device and equipment for a traffic monitoring system. Background Art

[0002] Existing traffic monitoring systems mainly focus on simple traffic violations such as running red lights, and mostly rely on single video monitoring data. It mainly obtains the video frame sequence to be detected through the camera, and combines the pre-trained vehicle detection model and signal light detection model to identify the vehicle's movement trajectory and the status of the signal light. The system determines whether the vehicle is in violation based on the relative position relationship between the vehicle trajectory, lane area and signal light status, and captures the violating vehicle according to the preset capture method when the vehicle is determined to be in violation. However, with the complexity of the traffic environment, traffic violations by traffic targets and unsafe driving behaviors of traffic targets (such as frequent lane changes and serpentine driving) are becoming more and more common, resulting in the following problems in the existing traffic monitoring system:

[0003] 1) Relying on the video frame sequence captured by the camera to detect traffic violations through a single data source, it is limited by external factors such as lighting conditions, obstructions and weather.

[0004] 2) The state of traffic lights depends on the accuracy of the detection model and is easily affected by occlusion and the accuracy of the detection model, so its accuracy and reliability are low.

[0005] 3) It focuses on detecting vehicle violations in a single lane and cannot accurately identify complex scenarios where vehicles evade red lights through multi-stage trajectories, such as Figure 6 As shown; it is also impossible to detect unsafe driving behaviors of the vehicle, such as Figure 7 Serpentine driving shown. Summary of the invention

[0006] The technical purpose to be achieved by the embodiments of the present application is to provide a driving behavior analysis method, device and equipment for a traffic monitoring system, so as to solve the problem that the existing traffic monitoring system has low accuracy in driving behavior recognition and is unable to recognize complex driving behaviors.

[0007] In order to solve the above technical problems, the embodiment of the present application provides a driving behavior analysis method of a traffic monitoring system, comprising:

[0008] Generate the trajectory points of traffic targets based on the perception data of multi-source sensors;

[0009] Matching the track points with lane lines on a high-precision map, and assigning lane attributes to the track points;

[0010] Associating traffic light status information with the track point, wherein the traffic light status information is determined based on detection data of a traffic signal detection model and received traffic police signal data;

[0011] When the number of the trajectory points reaches a preset number, generating a motion trajectory of the traffic target based on the trajectory points;

[0012] When the motion trajectory includes at least one inflection point, the motion trajectory is subjected to driving behavior recognition according to a preset multi-stage trajectory recognition strategy to determine the driving behavior of the traffic target, wherein each of the inflection points divides the motion trajectory into two stage trajectories, and the direction angle of the stage trajectories on both sides of the inflection point is greater than a first threshold.

[0013] Preferably, in the above method, the generating of the trajectory points of the traffic target based on the perception data of the multi-source sensors comprises:

[0014] Acquiring the perception data of the traffic target perceived by the multi-source sensors located in different directions;

[0015] Preprocessing the perception data to obtain the perception data in a unified coordinate system;

[0016] Based on a preset fusion algorithm, the preprocessed perception data is fused to generate the trajectory points.

[0017] Optionally, in the method described above, the multi-stage trajectory recognition strategy includes:

[0018] Not following the driving trajectory recognition strategy of the guided lane;

[0019] Avoiding red light running trajectory recognition strategy;

[0020] Snake driving trajectory recognition strategy.

[0021] Specifically, in the method as described above, when the multi-stage trajectory recognition strategy is the non-guided lane driving trajectory recognition strategy, performing driving behavior recognition on the motion trajectory includes:

[0022] Determining a first turning trend corresponding to the traffic target according to direction vectors corresponding to the trajectories of two adjacent stages;

[0023] The trajectory point in the trajectory determines the second turning trend corresponding to the traffic target according to the lane attribute corresponding to the trajectory point in the previous stage trajectory in two adjacent stage trajectories;

[0024] If the first turning trend is different from the second turning trend, it is determined that the traffic target has a driving behavior of not driving in the guide lane.

[0025] Preferably, in the above method, when the multi-stage trajectory recognition strategy is the straight-ahead red light avoidance trajectory recognition strategy, performing driving behavior recognition on the motion trajectory includes:

[0026] According to the traffic light status information corresponding to each trajectory point, a sub-motion trajectory within a red light cycle is intercepted from the motion trajectory, and a portion of the sub-motion trajectory is located at a traffic intersection;

[0027] If there are two inflection points in the sub-motion trajectory, and the stage trajectories corresponding to the two inflection points meet the preset conditions, it is determined that the traffic target has a driving behavior of avoiding running a red light by going straight;

[0028] The preset conditions include:

[0029] The first starting point of the first stage trajectory before the first turning point is located in the right turn lane;

[0030] The third turning trend corresponding to the first inflection point is a right turning trend;

[0031] The fourth turning trend corresponding to the second turning point is a left turning trend;

[0032] An angle between a first connecting line between a second end point of the second stage trajectory after the second inflection point and the first starting point and the first stage trajectory is less than a second threshold;

[0033] The component of the first connecting line in the first preset direction is greater than the component of the second connecting line in the first preset direction, the second connecting line is the connecting line between the second inflection point and the first starting point, and the first preset direction is the trend direction of the first starting point;

[0034] The component of the first connecting line in the second preset direction is smaller than the component of the second connecting line in the second preset direction, and the second preset direction is perpendicular to the first preset direction.

[0035] Preferably, in the above method, when the multi-stage trajectory recognition strategy is the serpentine driving trajectory recognition strategy, performing driving behavior recognition on the motion trajectory includes:

[0036] If within the second preset time period, there are multiple third turning points whose included angles between adjacent stage trajectories are greater than a third threshold, it is determined that the traffic target has serpentine driving behavior, and the third threshold is greater than the first threshold.

[0037] Optionally, the method as described above further comprises:

[0038] Acquire the motion trajectory corresponding to each of the traffic targets passing through the target area;

[0039] Determining the driving information of each of the traffic targets according to the motion trajectory;

[0040] The current first traffic congestion state of the target area is determined according to the driving information and / or a second traffic congestion state after at least a first preset time period is predicted.

[0041] Optionally, the method as described above further comprises:

[0042] When a traffic accident event is triggered, the movement trajectories of multiple target vehicles in a second preset time period before the accident occurs in the target area are obtained, wherein when it is detected that the trajectories of two traffic targets overlap or an input accident query instruction is received, it is determined that the traffic accident event is triggered;

[0043] An accident analysis report is generated according to the motion trajectory and the lane attributes, the traffic light status information, and the position information corresponding to each trajectory point in the motion trajectory.

[0044] Another embodiment of the present application further provides a control device, including:

[0045] A first processing module, for generating trajectory points of traffic targets based on perception data of multi-source sensors;

[0046] A second processing module is used to match the trajectory point with the lane line of the high-precision map and assign lane attributes to the trajectory point;

[0047] A third processing module, configured to associate traffic light status information with the track point, wherein the traffic light status information is determined based on detection data of a traffic signal detection model and received traffic police signal data;

[0048] A fourth processing module, configured to generate a motion trajectory of the traffic target based on the trajectory points when the number of the trajectory points reaches a preset number;

[0049] The fifth processing module is used to identify the driving behavior of the motion trajectory according to a preset multi-stage trajectory recognition strategy when the motion trajectory includes at least one inflection point, so as to determine the driving behavior of the traffic target, wherein each of the inflection points divides the motion trajectory into two stage trajectories, and the direction angle of the stage trajectories on both sides of the inflection point is greater than a first threshold.

[0050] Another embodiment of the present application also provides an electronic device, including a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the steps of the driving behavior analysis method of the traffic monitoring system as described above.

[0051] Yet another embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the driving behavior analysis method of the traffic monitoring system as described above are implemented.

[0052] Another embodiment of the present application further provides a computer program product, including computer instructions, which, when executed by a processor, implement the steps of the driving behavior analysis method of the traffic monitoring system as described above.

[0053] Compared with the prior art, the driving behavior analysis method, device and equipment of the traffic monitoring system provided in the embodiments of the present application have at least the following beneficial effects:

[0054] This application uses high-precision perception data fusion based on multi-source data fusion, which not only overcomes the limitations of a single sensor, but also solves problems such as occlusion, visual blind spots, environmental factors, accuracy, data redundancy and error detection, and provides more stable and accurate traffic target motion trajectories, especially dynamic tracking capabilities in complex traffic scenes. Segmenting the motion trajectory through inflection points and performing driving behavior recognition based on a multi-stage trajectory recognition strategy can effectively adapt to driving behavior detection in complex traffic environments, improve the traffic monitoring system's ability to detect violations in complex traffic scenes, and thus improve the efficiency and accuracy of traffic management. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 One of the flowcharts of the driving behavior analysis method of the traffic monitoring system of the present application;

[0056] Figure 2 The second flowchart of the driving behavior analysis method of the traffic monitoring system of the present application;

[0057] Figure 3 The third flowchart of the driving behavior analysis method of the traffic monitoring system of the present application;

[0058] Figure 4 The fourth flowchart of the driving behavior analysis method of the traffic monitoring system of the present application;

[0059] Figure 5 It is a structural schematic diagram of the control device of the present application;

[0060] Figure 6 A schematic diagram of the trajectory for traffic goals to avoid running a red light when driving straight;

[0061] Figure 7 It is a schematic diagram of the trajectory of the serpentine driving of the traffic target;

[0062] Figure 8 This is a schematic diagram of the standard intersection layout;

[0063] Fig. 9 A schematic diagram of the trajectory of a traffic target not traveling in a guide lane;

[0064] Fig.10 Trajectory decomposition diagram for avoiding red light running for straight driving for traffic goals. DETAILED DESCRIPTION

[0065] In order to make the technical problems, technical solutions and advantages to be solved by the application clearer, the following will be described in detail in conjunction with the accompanying drawings and specific embodiments. In the following description, specific details such as specific configurations and components are provided only to help fully understand the embodiments of the application. Therefore, it should be clear to those skilled in the art that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the application. In addition, for clarity and brevity, the description of known functions and structures has been omitted.

[0066] It should be understood that the references to "one embodiment" or "an embodiment" throughout the specification mean that the specific features, structures, or characteristics associated with the embodiment are included in at least one embodiment of the present application. Therefore, the references to "in one embodiment" or "in an embodiment" appearing throughout the specification do not necessarily refer to the same embodiment. In addition, these specific features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.

[0067] In the various embodiments of the present application, it should be understood that the size of the serial numbers of the following processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0068] It should be understood that the term "and / or" in this article is only a description of the association relationship of the associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0069] In the embodiments provided in the present application, it should be understood that "B corresponding to A" means that B is associated with A, and B can be determined according to A. However, it should also be understood that determining B according to A does not mean determining B only according to A, and B can also be determined according to A and / or other information.

[0070] See also Figure 1 An embodiment of the present application provides a driving behavior analysis method for a traffic monitoring system, comprising:

[0071] Step S101, based on the perception data of the multi-source sensors, the trajectory points of the traffic target are generated. In this step, the perception data includes but is not limited to the real-time location information, speed, direction, etc. of the traffic target; the perception devices corresponding to the multi-source sensors include but are not limited to cameras, radars, lidars and global positioning system (GPS) devices, wherein the camera provides image information and location information of the traffic target, the radar provides the distance, relative speed and location information of the traffic target and other objects, and the GPS device provides global positioning data of the traffic target to ensure the accurate positioning of the target.

[0072] Step S102, matching the trajectory point with the lane line of the high-precision map, and assigning lane attributes to the trajectory point. In this step, the location information (e.g., longitude and latitude information) of the trajectory point is matched with the lane line of the high-precision map to identify the lane attributes of the lane where the trajectory point is located (straight ahead, left turn, right turn, straight ahead and left turn, straight ahead and right turn, left turn and U-turn, etc.), and the lane attributes are assigned to the corresponding trajectory point to facilitate subsequent driving behavior analysis based on the lane attributes of the trajectory point.

[0073] Step S103, associating the traffic light status information with the trajectory point, wherein the traffic light status information is determined based on the detection data of the traffic signal detection model and the received traffic police signal data. In this step, in order to facilitate the subsequent driving behavior analysis based on the signal light status corresponding to the trajectory point (for example, determining whether the motion trajectory of the traffic target conforms to the current traffic light status), the traffic light status information is also obtained. Specifically, the traffic light status information is determined based on the detection data of the traffic signal detection model and the received traffic police signal data. Data verification and confidence superposition of the detection data and the traffic police signal data are beneficial to enhancing the accuracy of the traffic light status information and ensuring system reliability.

[0074] Step S104, when the number of the trajectory points reaches a preset number, the motion trajectory of the traffic target is generated based on the trajectory points. In this step, by setting the preset number, it can be ensured that there is sufficient data for analysis and judgment to ensure the accuracy of the generated motion trajectory. At the same time, it is ensured that there is sufficient data in the motion trajectory for driving behavior analysis. Optionally, due to the periodicity of data sampling, the standard for generating the motion trajectory can also be that the duration of detecting the traffic target reaches a preset time.

[0075] Step S105, when the motion trajectory includes at least one inflection point, the motion trajectory is subjected to driving behavior recognition according to a preset multi-stage trajectory recognition strategy to determine the driving behavior of the traffic target, wherein each inflection point divides the motion trajectory into two stage trajectories, and the direction angle of the stage trajectories on both sides of the inflection point is greater than a first threshold value. In this step, since the driving behavior judgment of the motion trajectory without inflection points is relatively simple, it is sufficient to judge whether there are driving behaviors such as speeding and running a red light based on the speed, position and traffic light status of the traffic target, so no specific description is given in this application. This embodiment mainly recognizes the complex driving behavior corresponding to the motion trajectory with at least one inflection point. At this time, since the motion trajectory has multiple inflection points, the motion trajectory can be divided into multiple stage trajectories based on the inflection points, and then the motion trajectory composed of multiple stage trajectories is subjected to driving behavior recognition according to the preset multi-stage trajectory recognition strategy to determine the driving behavior of the traffic target.

[0076] In summary, this application performs perception data fusion based on high precision of multi-source data fusion, which not only overcomes the limitations of a single sensor, but also solves problems such as occlusion, visual blind spots, environmental factors, accuracy, data redundancy and error detection, and provides more stable and accurate traffic target motion trajectories, especially dynamic tracking capabilities in complex traffic scenes. Segmenting the motion trajectory through inflection points and performing driving behavior recognition based on a multi-stage trajectory recognition strategy can effectively adapt to driving behavior detection in complex traffic environments, improve the traffic monitoring system's ability to detect violations in complex traffic scenes, and thus improve the efficiency and accuracy of traffic management.

[0077] It should be noted that the multi-source sensor described in this application is used to indicate that the sources of the perception data have multiple types, such as video data, radar data, and GPS data, and is not limited to all perception data being obtained through one device or multiple devices. For example, a camera and a radar can be combined into one device, or each can be a separate device.

[0078] See also Figure 2 Preferably, in the above method, the generating of the trajectory points of the traffic target based on the perception data of the multi-source sensors comprises:

[0079] Step S201, acquiring the perception data of the traffic target perceived by the multi-source sensors located in different directions;

[0080] Step S202, preprocessing the sensed data to obtain the sensed data in a unified coordinate system;

[0081] Step S203: Based on a preset fusion algorithm, the preprocessed perception data is fused to generate the trajectory points.

[0082] In this embodiment, since the traffic intersection has intersections in multiple directions, the multi-source sensor can be set at each intersection, such as the crossroads (such as Figure 8 As shown in the figure, multi-source sensors can be set at intersections in four directions. When acquiring data, perception data of traffic targets perceived by multi-source sensors in different directions can be acquired, and the perception data can be processed into data in a unified coordinate system through preprocessing means, wherein the preprocessing means include but are not limited to time and space alignment of the perception data. The unified coordinate system is preferably a standard coordinate system constructed based on the intersection direction of the traffic intersection. Taking a standard intersection as an example, the y-axis of the constructed standard coordinate system is from south to north, and the x-axis is from west to east. Then, based on a preset fusion algorithm (for example: Kalman filtering and / or Hungarian matching, etc.), each perception data can be fused to obtain trajectory points with high accuracy.

[0083] Optionally, in the method described above, the multi-stage trajectory recognition strategy includes:

[0084] Not following the driving trajectory recognition strategy of the guided lane;

[0085] Avoiding red light running trajectory recognition strategy;

[0086] Snake driving trajectory recognition strategy.

[0087] In this embodiment, the multi-stage trajectory recognition strategy corresponding to some illegal driving behaviors and unsafe driving behaviors is used as an example, which does not mean that this application is only used to identify the above three illegal driving behaviors and unsafe driving behaviors. To facilitate the understanding of those skilled in the art, the specific identification steps of the above illegal driving behaviors or unsafe driving behaviors are exemplified below.

[0088] Specifically, in the method as described above, when the multi-stage trajectory recognition strategy is the non-guided lane driving trajectory recognition strategy, performing driving behavior recognition on the motion trajectory includes:

[0089] Determine a first turning trend corresponding to the traffic target according to direction vectors corresponding to two adjacent stage trajectories, wherein the direction vector of each stage trajectory is a vector from the starting point to the end point of the stage trajectory;

[0090] Determining a second turning trend corresponding to the traffic target according to the lane attributes corresponding to the trajectory points in the previous stage trajectory in two adjacent stage trajectories;

[0091] If the first turning trend is different from the second turning trend, it is determined that the traffic target has a driving behavior of not driving in the guide lane.

[0092] In this embodiment, the schematic diagram of the trajectory of the traffic target not traveling along the guide lane is as follows: Fig. 9 As shown, when judging the driving behavior of not driving in the guide lane, the first turning trend corresponding to the traffic target is determined based on the direction vectors corresponding to the two adjacent stage trajectories, and the second turning trend corresponding to the traffic target is determined based on the lane attributes corresponding to the trajectory points in the previous stage trajectory of the two stage trajectories. If the first turning trend and the second turning trend are different, it can be determined that the traffic target has a driving behavior of not driving in the guide lane.

[0093] In a specific embodiment, if according to the direction vectors corresponding to the trajectories of two adjacent stages, it is determined that the first turning trend corresponding to the traffic target is a left turn, and the lane attribute corresponding to the trajectory point in the previous stage trajectory is straight driving, and it is determined that the traffic target has a driving behavior of not driving in the guide lane.

[0094] Preferably, in the above method, when the multi-stage trajectory recognition strategy is the straight-ahead red light avoidance trajectory recognition strategy, performing driving behavior recognition on the motion trajectory includes:

[0095] According to the traffic light status information corresponding to each trajectory point, a sub-motion trajectory within a red light cycle is intercepted from the motion trajectory, and a portion of the sub-motion trajectory is located at a traffic intersection;

[0096] If there are two inflection points in the sub-motion trajectory, and the stage trajectories corresponding to the two inflection points meet the preset conditions, it is determined that the traffic target has a driving behavior of avoiding running a red light by going straight;

[0097] The preset conditions include:

[0098] The first starting point of the first stage trajectory before the first turning point is located in the right turn lane;

[0099] The third turning trend corresponding to the first inflection point is a right turning trend;

[0100] The fourth turning trend corresponding to the second turning point is a left turning trend;

[0101] An angle between a first connecting line between a second end point of the second stage trajectory after the second inflection point and the first starting point and the first stage trajectory is less than a second threshold;

[0102] The component of the first connecting line in the first preset direction is greater than the component of the second connecting line in the first preset direction, the second connecting line is the connecting line between the second inflection point and the first starting point, and the first preset direction is the trend direction of the first starting point;

[0103] The component of the first connecting line in the second preset direction is smaller than the component of the second connecting line in the second preset direction, and the second preset direction is perpendicular to the first preset direction.

[0104] In this embodiment, the schematic diagram of the trajectory of the traffic target avoiding running a red light by going straight is as follows: Figure 6 As shown, when judging whether the traffic target has a driving behavior of avoiding a straight red light, since it is actually a red light running behavior, the traffic light status information corresponding to each trajectory point is selected, and a sub-motion trajectory within a red light cycle is intercepted from the motion trajectory, and the sub-motion trajectory needs to be partially located at the traffic intersection. At this time, based on the number of turning points in the sub-motion trajectory and the stage trajectory corresponding to the turning point, it is judged whether the preset conditions are met, wherein the setting of the preset conditions is used to judge whether the traffic target avoids the straight red light by turning right first and then turning left in the right turn lane. Among them, the angle between the first connecting line and the first stage trajectory is used to judge whether the lane has been changed, and the comparison of the components of the first connecting line and the second connecting line is used to further judge whether the final direction of the traffic target is straight.

[0105] In a specific embodiment, based on Fig.10 The trajectory decomposition diagram of the traffic target avoiding running a red light while driving straight ahead is shown. In a red light cycle, the traffic target is located in the right turn lane at the first starting point S of the first stage trajectory. The first turning point A at the traffic intersection is a right turn trend, followed by a second turning point B with a left turn trend, which is a left turn trend. At this time, there is a second end point C of the second stage trajectory, and the angle between the trend vector y of SC and the first starting point S is less than the second set threshold value θ, and the length of the x-axis component vector SCx of the SC vector is less than the length of the x-axis component vector SBx of the SB vector, and the length of the y-axis component vector SCx of the SC vector is greater than the length of the y-axis component vector SBx of the SB vector. Then, it is judged that the behavior corresponding to the trajectory avoids driving through a red light.

[0106] Preferably, in the above method, when the multi-stage trajectory recognition strategy is the serpentine driving trajectory recognition strategy, performing driving behavior recognition on the motion trajectory includes:

[0107] If within the second preset time period, there are multiple third turning points whose included angles between adjacent stage trajectories are greater than a third threshold, it is determined that the traffic target has serpentine driving behavior, and the third threshold is greater than the first threshold.

[0108] In this embodiment, the trajectory diagram of serpentine driving is as follows: Fig. 9 As shown, it can judge whether the traffic target is a serpentine driving behavior based on the number of sharp turning points in a shorter time (a second preset time period), where the angle between the adjacent stage trajectories corresponding to the sharp turning point, i.e., a turning point, is greater than a third threshold, and the third threshold is greater than the first threshold.

[0109] See also Figure 3 Optionally, the method as described above further comprises:

[0110] Step S301, obtaining the motion trajectory corresponding to each of the traffic targets passing through the target area;

[0111] Step S302, determining the travel information of each of the traffic targets according to the movement trajectory;

[0112] Step S303: determining a current first traffic congestion state of the target area according to the driving information and / or predicting a second traffic congestion state at least after a first preset time period.

[0113] In this embodiment, when performing the analysis, the driving information (such as driving speed, dwell time, position change, etc.) of each traffic target is obtained based on the motion trajectory corresponding to each traffic target passing through the same target area (including the position coordinates, speed, time, etc. corresponding to each trajectory point), and based on the driving information, it is determined whether the target area is currently congested or whether congestion will occur in the future, so as to realize the assessment of the road traffic congestion situation.

[0114] See also Figure 4 Optionally, the method as described above further comprises:

[0115] Step S401, when a traffic accident event is triggered, obtaining the movement trajectories of multiple target vehicles in a second preset time period before the accident occurs in the target area, wherein when it is detected that the trajectories of two traffic targets overlap or an input accident query instruction is received, it is determined that the traffic accident event is triggered;

[0116] Step S402: Generate an accident analysis report based on the motion trajectory and the lane attributes, traffic light status information, and location information corresponding to each trajectory point in the motion trajectory.

[0117] In this embodiment, when performing analysis, if a traffic accident event is triggered based on detection or user input, the motion trajectories of multiple target vehicles within a second preset time period before the traffic accident event occurs in the target area are obtained, and an accident analysis report is generated based on each motion trajectory and the lane attributes, traffic light status information, and location information corresponding to each trajectory point in the motion trajectory. The accident analysis report is used to illustrate the motion trajectory of each relevant traffic target, as well as the driving behavior obtained based on the motion trajectory (for example: speeding, running a red light, etc.), thereby providing data support for analyzing the cause of the accident, determining responsibility, etc.

[0118] See also Figure 5 Another embodiment of the present application further provides a control device, comprising:

[0119] The first processing module 501 is used to generate trajectory points of traffic targets based on the perception data of multi-source sensors;

[0120] The second processing module 502 is used to match the trajectory point with the lane line of the high-precision map and assign lane attributes to the trajectory point;

[0121] A third processing module 503 is used to associate traffic light status information with the track point, wherein the traffic light status information is determined according to detection data of a traffic signal detection model and received traffic police signal data;

[0122] A fourth processing module 504 is configured to generate a motion trajectory of the traffic target based on the trajectory points when the number of the trajectory points reaches a preset number;

[0123] The fifth processing module 505 is used to identify the driving behavior of the motion trajectory and determine the driving behavior of the traffic target according to at least one preset multi-stage trajectory recognition strategy when the motion trajectory includes at least one inflection point, wherein each of the inflection points divides the motion trajectory into two stage trajectories, and the direction angle of the stage trajectories on both sides of the inflection point is greater than the first threshold.

[0124] Preferably, in the control device as described above, the first processing module comprises:

[0125] A first processing unit, configured to obtain the perception data of the traffic target perceived by the multi-source sensors located in different directions;

[0126] A second processing unit, configured to pre-process the perception data to obtain the perception data in a unified coordinate system;

[0127] The third processing unit is used to perform fusion processing on the preprocessed perception data based on a preset fusion algorithm to generate the trajectory points.

[0128] Optionally, in the control device as described above, the multi-stage trajectory recognition strategy includes:

[0129] Not following the driving trajectory recognition strategy of the guided lane;

[0130] Avoiding red light running trajectory recognition strategy;

[0131] Snake driving trajectory recognition strategy.

[0132] Specifically, in the control device as described above, when the multi-stage trajectory recognition strategy is the non-guided lane driving trajectory recognition strategy, the fifth processing module includes:

[0133] A fourth processing unit, configured to determine a first turning trend corresponding to the traffic target according to direction vectors corresponding to the trajectories of two adjacent stages;

[0134] A fifth processing unit, configured to determine a second turning trend corresponding to the traffic target according to the lane attribute corresponding to the trajectory point in the previous stage trajectory in two adjacent stage trajectories;

[0135] The sixth processing unit is used to determine that the traffic target has a driving behavior of not driving in the guide lane if the first turning trend and the second turning trend are different.

[0136] Preferably, in the control device as described above, when the multi-stage trajectory recognition strategy is the straight-ahead red light avoidance trajectory recognition strategy, the fifth processing module includes:

[0137] a seventh processing unit, configured to intercept a sub-motion trajectory within a red light cycle from the motion trajectory according to the traffic light state information corresponding to each trajectory point, wherein a portion of the sub-motion trajectory is located at a traffic intersection;

[0138] an eighth processing unit, configured to determine that the traffic target has a driving behavior of avoiding running a red light by going straight if there are two inflection points in the sub-motion trajectory and the stage trajectories corresponding to the two inflection points meet a preset condition;

[0139] The preset conditions include:

[0140] The first starting point of the first stage trajectory before the first turning point is located in the right turn lane;

[0141] The third turning trend corresponding to the first inflection point is a right turning trend;

[0142] The fourth turning trend corresponding to the second turning point is a left turning trend;

[0143] An angle between a first connecting line between a second end point of the second stage trajectory after the second inflection point and the first starting point and the first stage trajectory is less than a second threshold;

[0144] The component of the first connecting line in the first preset direction is greater than the component of the second connecting line in the first preset direction, the second connecting line is the connecting line between the second inflection point and the first starting point, and the first preset direction is the trend direction of the first starting point;

[0145] The component of the first connecting line in the second preset direction is smaller than the component of the second connecting line in the second preset direction, and the second preset direction is perpendicular to the first preset direction.

[0146] Preferably, in the control device as described above, when the multi-stage trajectory recognition strategy is the serpentine driving trajectory recognition strategy, the fifth processing module includes:

[0147] The ninth processing unit is used to determine that the traffic target has serpentine driving behavior if, within a second preset time period, there are multiple third turning points whose included angles between adjacent stage trajectories are greater than a third threshold, and the third threshold is greater than the first threshold.

[0148] Optionally, the control device as described above further includes:

[0149] A sixth processing module, used for obtaining the motion trajectory corresponding to each of the traffic targets passing through the target area;

[0150] A seventh processing module, used to determine the travel information of each of the traffic targets according to the motion trajectory;

[0151] An eighth processing module is used to determine a current first traffic congestion state of the target area according to the driving information and / or predict a second traffic congestion state at least after a first preset time period.

[0152] Optionally, the control device as described above further includes:

[0153] A ninth processing module, configured to obtain the movement trajectories of a plurality of target vehicles within a second preset time period before the accident occurs in the target area when a traffic accident event is triggered, wherein when it is detected that the trajectories of two traffic targets overlap or an input accident query instruction is received, it is determined that the traffic accident event is triggered;

[0154] The tenth processing module is used to generate an accident analysis report according to the motion trajectory and the lane attributes corresponding to each trajectory point in the motion trajectory, the traffic light status information, and the location information.

[0155] The control device embodiment of the present application is a device corresponding to the embodiment of the driving behavior analysis method of the above-mentioned traffic monitoring system. All the implementation means in the above-mentioned method embodiment are applicable to the embodiment of the device and can achieve the same technical effect. The above-mentioned control device provided in the embodiment of the present application can implement all the method steps implemented in the above-mentioned method embodiment and can achieve the same technical effect. The parts and beneficial effects that are the same as those in the method embodiment in this embodiment will not be specifically described here.

[0156] Another embodiment of the present application further provides an electronic device, including a processor, a memory, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, the steps of the driving behavior analysis method of the traffic monitoring system as described above are implemented, which will not be described again here to avoid repetition.

[0157] Another embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the driving behavior analysis method of the traffic monitoring system as described above are implemented. To avoid repetition, they are not described here.

[0158] Another embodiment of the present application further provides a computer program product, including computer instructions, which, when executed by a processor, implement the steps of the driving behavior analysis method of the traffic monitoring system as described above, which will not be described again here to avoid repetition.

[0159] In addition, the present application may repeat reference numerals and / or letters in different examples. This repetition is for the purpose of simplicity and clarity, and does not in itself indicate the relationship between the various embodiments and / or settings discussed.

[0160] It should also be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusions.

[0161] The above is a preferred embodiment of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles described in the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A driving behavior analysis method for a traffic monitoring system, characterized in that: include: Generate the trajectory points of traffic targets based on the perception data of multi-source sensors; Matching the track points with lane lines on a high-precision map, and assigning lane attributes to the track points; Associating traffic light status information with the track point, wherein the traffic light status information is determined based on detection data of a traffic signal detection model and received traffic police signal data; When the number of the trajectory points reaches a preset number, generating a motion trajectory of the traffic target based on the trajectory points; When the motion trajectory includes at least one inflection point, driving behavior recognition is performed on the motion trajectory according to at least one preset multi-stage trajectory recognition strategy to determine the driving behavior of the traffic target, wherein each of the inflection points divides the motion trajectory into two stage trajectories, and the direction angle of the stage trajectories on both sides of the inflection point is greater than a first threshold.

2. The method according to claim 1, characterized in that The generating of the trajectory points of the traffic target based on the perception data of the multi-source sensors includes: Acquiring the perception data of the traffic target perceived by the multi-source sensors located in different directions; Preprocessing the perception data to obtain the perception data in a unified coordinate system; Based on a preset fusion algorithm, the preprocessed perception data is fused to generate the trajectory points.

3. The method according to claim 1, characterized in that The multi-stage trajectory recognition strategy includes: Not following the driving trajectory recognition strategy of the guided lane; Avoiding red light running trajectory recognition strategy; Snake driving trajectory recognition strategy.

4. The method according to claim 3, characterized in that When the multi-stage trajectory recognition strategy is the non-guided lane driving trajectory recognition strategy, performing driving behavior recognition on the motion trajectory includes: Determining a first turning trend corresponding to the traffic target according to direction vectors corresponding to the trajectories of two adjacent stages; The trajectory point in the trajectory determines the second turning trend corresponding to the traffic target according to the lane attribute corresponding to the trajectory point in the previous stage trajectory in two adjacent stage trajectories; If the first turning trend is different from the second turning trend, it is determined that the traffic target has a driving behavior of not driving in the guide lane.

5. The method according to claim 3, characterized in that: When the multi-stage trajectory recognition strategy is the straight-ahead red light avoidance trajectory recognition strategy, performing driving behavior recognition on the motion trajectory includes: According to the traffic light status information corresponding to each trajectory point, a sub-motion trajectory within a red light cycle is intercepted from the motion trajectory, and a portion of the sub-motion trajectory is located at a traffic intersection; If there are two inflection points in the sub-motion trajectory, and the stage trajectories corresponding to the two inflection points meet the preset conditions, it is determined that the traffic target has a driving behavior of avoiding running a red light by going straight; The preset conditions include: The first starting point of the first stage trajectory before the first turning point is located in the right turn lane; The third turning trend corresponding to the first inflection point is a right turning trend; The fourth turning trend corresponding to the second turning point is a left turning trend; An angle between a first connecting line between a second end point of the second stage trajectory after the second inflection point and the first starting point and the first stage trajectory is less than a second threshold; The component of the first connecting line in the first preset direction is greater than the component of the second connecting line in the first preset direction, the second connecting line is the connecting line between the second inflection point and the first starting point, and the first preset direction is the trend direction of the first starting point; The component of the first connecting line in the second preset direction is smaller than the component of the second connecting line in the second preset direction, and the second preset direction is perpendicular to the first preset direction.

6. The method according to claim 3, characterized in that When the multi-stage trajectory recognition strategy is the serpentine driving trajectory recognition strategy, performing driving behavior recognition on the motion trajectory includes: If within the second preset time period, there are multiple third turning points whose included angles between adjacent stage trajectories are greater than a third threshold, it is determined that the traffic target has serpentine driving behavior, and the third threshold is greater than the first threshold.

7. The method according to claim 1, characterized in that Also includes: Acquire the motion trajectory corresponding to each of the traffic targets passing through the target area; Determining the driving information of each of the traffic targets according to the motion trajectory; The current first traffic congestion state of the target area is determined according to the driving information and / or a second traffic congestion state after at least a first preset time period is predicted.

8. The method according to claim 1, characterized in that Also includes: When a traffic accident event is triggered, the movement trajectories of multiple target vehicles in a second preset time period before the accident occurs in the target area are obtained, wherein when it is detected that the trajectories of two traffic targets overlap or an input accident query instruction is received, it is determined that the traffic accident event is triggered; An accident analysis report is generated according to the motion trajectory and the lane attributes, the traffic light status information, and the position information corresponding to each trajectory point in the motion trajectory.

9. A control device, characterized in that: include: A first processing module, for generating trajectory points of traffic targets based on perception data of multi-source sensors; A second processing module is used to match the trajectory point with the lane line of the high-precision map and assign lane attributes to the trajectory point; A third processing module, configured to associate traffic light status information with the track point, wherein the traffic light status information is determined based on detection data of a traffic signal detection model and received traffic police signal data; A fourth processing module, configured to generate a motion trajectory of the traffic target based on the trajectory points when the number of the trajectory points reaches a preset number; The fifth processing module is used to identify the driving behavior of the motion trajectory according to a preset multi-stage trajectory recognition strategy when the motion trajectory includes at least one inflection point, so as to determine the driving behavior of the traffic target, wherein each of the inflection points divides the motion trajectory into two stage trajectories, and the direction angle of the stage trajectories on both sides of the inflection point is greater than a first threshold.

10. An electronic device, characterized in that: The invention comprises a processor, a memory and a computer program stored in the memory and executable on the processor, wherein when the computer program is executed by the processor, the steps of the driving behavior analysis method of the traffic monitoring system as described in any one of claims 1 to 8 are implemented.

11. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the driving behavior analysis method of the traffic monitoring system according to any one of claims 1 to 8 are implemented.

12. A computer program product, characterized in that The method comprises computer instructions, which, when executed by a processor, implement the steps of the driving behavior analysis method of the traffic monitoring system according to any one of claims 1 to 8.

Images