Traffic collision accident detection method, device and readable medium

Abstract

The present application provides a traffic collision accident detection method, device and readable medium. The present application first obtains multiple road perception data of the road section to be detected; then determines the current collision accident type based on the intensity of the sound; finally, based on the current collision accident type, combined with the multiple road perception data, determines the accident information of the traffic collision accident. By coordinating the road perception data of multiple sensors, first determining the collision type based on the sound, and then combining all the road perception data to comprehensively judge whether it is a collision accident, the application solves the drawbacks of traffic flow comparison-based and video-based detection, and provides road traffic management departments with richer accident information to facilitate timely processing.

Description

Technical Field

[0001] The present application relates to the field of traffic detection, and in particular to a method, device and readable medium for detecting traffic collision accidents. Background Art

[0002] Traffic accident detection and identification is a key research topic in road traffic management. Efficient and accurate traffic accident identification technology can provide decision-making support for subsequent emergency management and rescue efforts. Among the many types of traffic anomalies, vehicle collisions often signal the occurrence of a traffic accident. The ability to respond promptly and implement appropriate rescue, evacuation, and isolation measures after an accident significantly impacts the survival rate of victims and traffic conditions at the accident site. Therefore, the ability to automatically, quickly, and accurately detect collisions in traffic surveillance video is an essential core capability of intelligent transportation systems.

[0003] Generally speaking, existing traffic accident identification solutions mainly include two modes: active reporting of traffic accident information and data-driven traffic accident identification. The traffic accident information reporting mode suffers from low reporting efficiency, incomplete reporting, resource consumption, and is prone to underreporting and manual data entry errors. Data-driven methods can be divided into two categories: indirect detection by comparing traffic flow over a period of time with historical normal traffic flow, and active detection and identification through visual cameras. Data-driven methods are subject to environmental and road conditions, resulting in large detection errors and high difficulty, and thus have many shortcomings. Summary of the Invention

[0004] In response to the problems existing in the traffic accident identification solutions in the prior art, the present application provides a traffic collision accident detection method, device and readable medium.

[0005] A first embodiment of the present application provides a method for detecting a traffic collision accident, comprising:

[0006] Acquire multiple types of road perception data for the road section to be detected, where the types of the road perception data include sound, and each type of road perception data is collected by a corresponding sensor;

[0007] determining the type of the current collision accident according to the intensity of the sound;

[0008] According to the current collision accident type, combined with the multiple road perception data, accident information of the traffic collision accident is determined.

[0009] In an optional embodiment, the collision accident type includes a minor collision and a severe collision. Correspondingly, determining the current collision accident type according to the intensity of the sound includes:

[0010] If the intensity of the collected sound is higher than a set threshold, the current collision accident type is determined to be a severe collision, otherwise it is a minor collision.

[0011] In an optional embodiment, the type of road perception data further includes: vehicle motion data, the corresponding sensor including at least one of a lidar and a camera;

[0012] The determining of the accident information of the traffic collision accident based on the current collision accident type and in combination with the multiple road perception data includes:

[0013] If the current collision type is a severe collision, estimating an estimated azimuth angle and an estimated distance of the current collision relative to the sensor based on the sound data;

[0014] Accident information of the traffic collision accident is determined based on the estimated azimuth and the estimated distance in combination with the vehicle motion data within a corresponding position range.

[0015] In an optional embodiment, the type of road perception data further includes: vehicle motion data, the corresponding sensor including at least one of a lidar and a camera;

[0016] The determining of the accident information of the traffic collision accident based on the current collision accident type and in combination with the multiple road perception data includes:

[0017] If the current collision type is a minor collision, determining a suspected traffic collision accident based on all vehicle motion data in the entire road section;

[0018] determining whether a traffic collision has occurred based on the sound data within the location range of each suspected traffic collision;

[0019] If a traffic collision is determined to have occurred, the point cloud data collected by the lidar within the corresponding position range, as well as the vehicle images of the relevant vehicles in the camera images and the accident video are retrieved.

[0020] In an optional embodiment, the sound source sensors are distributed at multiple monitoring sites, and estimating the estimated azimuth and estimated distance of the current collision relative to the sensors based on the sound data includes:

[0021] Based on the time delay estimation algorithm, the distance between the sound source and each sound source sensor is determined with one of the sound source sensors as the reference origin;

[0022] Calculating, based on the distance between the sound source and each sound source sensor, a first distance between the sound source and the i-th sound source sensor and a second distance between the sound source and the j-th sound source sensor, and determining a distance difference between the first distance and the second distance; i and j are different and both less than N;

[0023] An estimated azimuth and an estimated distance of the current collision relative to the sensor are determined based on the speed of sound and each distance difference.

[0024] In an optional embodiment, the vehicle motion data includes: speed, position, heading angle, and distance between adjacent vehicles at each moment; the accident information includes vehicle images, accident videos, and point cloud data; correspondingly, the sensors include lidar and cameras;

[0025] The determining of the accident information of the traffic collision accident based on the estimated azimuth and the estimated distance in combination with the vehicle motion data within the corresponding position range includes:

[0026] If, within a position range corresponding to the azimuth and distance of the current collision relative to the sensor, the change in the speed, position, heading angle, and distance between adjacent vehicles within a set time period exceeds a corresponding threshold, it is determined that a traffic collision accident has occurred within the corresponding position range;

[0027] Retrieve the point cloud data collected by the lidar within the corresponding position range, as well as the vehicle image and accident video of the relevant vehicle in the camera image.

[0028] In an optional embodiment, the vehicle motion data includes: speed, position, heading angle, and distance between adjacent vehicles; and determining whether a traffic collision has occurred based on the sound data within the location range of each suspected traffic collision accident includes:

[0029] Determine the location range of a suspected traffic collision based on the distance between adjacent vehicles in the entire road section at each moment, combined with the speed, position, and heading angle of each vehicle;

[0030] Within the location range of the suspected traffic collision accident, it is determined whether the suspected traffic collision accident is a misjudgment based on the sound data in the location range. If not, it is determined that a traffic collision accident has occurred within the location range.

[0031] In an optional embodiment, determining the location range of the suspected traffic collision accident based on the distance between adjacent vehicles in the entire road section at each moment in combination with the speed, position, and heading angle of each vehicle includes:

[0032] Splitting the road section to be detected into multiple sub-sections, and obtaining a historical vehicle distance dataset of the multiple sub-sections;

[0033] For each sub-segment, the change in the distance between adjacent vehicles in the current period within a set time period is analyzed for correlation with the change in the distance between adjacent vehicles in the historical vehicle distance dataset for the sub-segment when there are no accidents and when there are accidents, to determine whether the change in the current sub-segment is abnormal.

[0034] If there is an abnormality, the speed, position and heading angle of each vehicle are combined to determine whether the consistency of the speed change, position change and heading angle change of each vehicle is higher than the set threshold;

[0035] If so, the corresponding sub-segment is determined as the location range of the suspected traffic collision accident.

[0036] A second embodiment of the present application provides a traffic collision accident detection device, comprising:

[0037] an acquisition module for acquiring a variety of road perception data of the road section to be detected, wherein the types of the road perception data include sound, and each type of road perception data is collected by a corresponding sensor;

[0038] A type determination module, which determines the type of the current collision accident based on the intensity of the sound;

[0039] The accident information determination module determines the accident information of the traffic collision accident based on the current collision accident type and in combination with the multiple road perception data.

[0040] A third aspect of the present application provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method described above when executing the computer program.

[0041] A fourth aspect of the present application provides a computer-readable storage medium having a computer program stored thereon, which implements the steps of the method described above when the computer program is executed by a processor.

[0042] It can be seen from the above technical solution that the present application provides a traffic collision accident detection method, device and readable medium. The present application coordinates the road perception data of multiple sensors, and then first determines the collision type based on the sound, and then combines all the road perception data to comprehensively judge whether it is a collision accident. It solves the drawbacks of traffic flow comparison and video detection, and provides road traffic management departments with richer accident information to facilitate timely processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0044] Figure 1 It is a flow chart of the traffic collision accident detection method in an embodiment of the present application.

[0045] Figure 2 This is one of the detailed flow charts of the sub-steps of the traffic collision accident detection method in the embodiment of the present application.

[0046] Figure 3 This is the second detailed flow chart of the sub-steps of the traffic collision accident detection method in the embodiment of the present application.

[0047] Figure 4 This is the third detailed flow chart of the sub-steps of the traffic collision accident detection method in the embodiment of the present application.

[0048] Figure 5 This is the fourth detailed flow chart of the sub-steps of the traffic collision accident detection method in the embodiment of the present application.

[0049] Figure 6 It is a structural diagram of the traffic collision accident detection device in an embodiment of the present application.

[0050] Figure 7 It is a schematic diagram of the specific structure of the electronic device in the embodiment of the present application. DETAILED DESCRIPTION

[0051] To make the purpose, technical solutions, and advantages of the embodiments of this application more clear, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0052] It should be noted that the traffic collision accident detection method, system, electronic device and computer-readable storage medium disclosed in this application can be used in the field of traffic collision accident detection technology, and can also be used in any field other than the field of traffic collision accident detection technology. The application field of the traffic collision accident detection method, system, electronic device and computer-readable storage medium disclosed in this application is not limited.

[0053] Traffic accident detection and identification is a key research topic in road traffic management. Efficient and accurate traffic accident identification technology can provide decision-making support for subsequent emergency management and rescue efforts. Among the many types of traffic anomalies, vehicle collisions often signal the occurrence of a traffic accident. The ability to respond promptly and implement appropriate rescue, evacuation, and isolation measures after an accident significantly impacts the survival rate of victims and traffic conditions at the accident site. Therefore, the ability to automatically, quickly, and accurately detect collisions in traffic surveillance video is an essential core capability of intelligent transportation systems.

[0054] Generally speaking, existing traffic accident identification solutions mainly include two modes: active reporting of traffic accident information and data-driven traffic accident identification. Among them, the traffic accident information reporting mode has low reporting efficiency, incomplete reporting information, wastes human resources, is prone to omissions, and is prone to manual input errors.

[0055] Data-driven detection can be categorized into two main types: indirect detection by comparing traffic flow over a period of time with historical normal traffic flow patterns, and active accident detection and identification using visual cameras. Traffic flow analysis: When a traffic accident occurs on a certain road section, traffic flow parameters on the current section and upstream and downstream sections will change significantly: traffic will queue upstream of the accident area or flow at a higher density, while vehicles on the upstream section will slow down, creating a significant speed difference with vehicles approaching the upstream section.

[0056] However, the traffic flow comparison method primarily relies on changes in traffic flow parameters on the target road segment and its upstream and downstream sections. However, other factors besides traffic accidents can cause similar changes, such as changes in traffic demand and temporary traffic control measures. Therefore, data-driven traffic accident identification models can be affected by these factors, leading to false reports. Furthermore, the response time is long, and traffic accidents may not be detected immediately.

[0057] Video-based collision detection methods use visual detection to determine whether a vehicle has collided. However, visual detection has low positioning accuracy, and the estimated distance between vehicles can be subject to significant errors, making false alarms more likely. Furthermore, traffic collisions occur very quickly, making it difficult to detect abnormal distances. Furthermore, visual detection is susceptible to environmental influences, such as darkness, strong sunlight, rain, snow, and fog, making it difficult to detect all-weather detection.

[0058] Based on this, the present application provides an implementation method of a traffic collision accident detection method, such as Figure 1 Shown, including:

[0059] S1: Acquire a variety of road perception data of a road section to be detected, where the types of the road perception data include sound, and each type of road perception data is collected by a corresponding sensor.

[0060] S2: Determine the type of the current collision accident according to the intensity of the sound.

[0061] S3: Determine accident information of the traffic collision accident based on the current collision accident type and in combination with the multiple road perception data.

[0062] From the above description, it can be seen that the traffic collision accident detection method provided in the embodiment of the present application coordinates the road perception data of multiple sensors, and then first determines the collision type based on the sound, and then combines all the road perception data to comprehensively judge whether it is a collision accident. It solves the drawbacks of traffic flow comparison and video detection, and provides road traffic management departments with richer accident information to facilitate timely processing.

[0063] In the embodiment of the present application, the perception data is secondary data obtained by computer processing after being collected by the sensor. For example, the perception data of the lidar is the vehicle's speed, distance, position and heading angle, etc. The perception data of the visual camera is visual information such as the vehicle's color, model, license plate number, etc. The perception data of the sound source sensor is the sound within the detection range.

[0064] It can be understood that the laser radar receives the echo signal through the detector as primary data, and then processes the echo signal to form point cloud data, and analyzes the point cloud data (for example, continuous wave frequency modulation ranging) to obtain secondary data. Similarly, the vehicle's color, model, license plate number and other visual information are secondary data formed by computer recognition of the image taken by the visual camera. This application will not go into details about this.

[0065] It can be understood that road perception data includes sound data and the vehicle motion data formed by processing point cloud data, images and video data as mentioned above, that is, data related to vehicle driving such as vehicle speed and vehicle spacing. Furthermore, road perception data also includes vehicle attribute data, such as vehicle model, license plate number, vehicle sound color, etc., as well as vehicle road status data, such as vehicle position, vehicle heading angle, etc.

[0066] In a possible implementation, the sensor may include a lidar, a visual camera, and a sound source detector, etc. Of course, the present application is not limited to this. The perception data collected by the lidar is point cloud data, the data collected by the visual camera is image data or video data, and the data collected by the sound source detector is sound data. The present application does not elaborate on this. It can be understood that the present application can also be configured with other sensors, such as infrared sensors, to use infrared light to assist in ranging or to assist in determining whether the vehicle changes lanes.

[0067] For example, an infrared sensor is set on the lane line of a traffic road. When a vehicle changes lanes on this road section, it will trigger a change in the transmission and reception of infrared signals, thereby determining that the vehicle has changed lanes. This application does not go into details about this.

[0068] LiDAR includes a main control chip, a scanning module, a transmitting module, and a receiving module. The scanning module generally includes a scanner and scanner driver, the transmitting module includes a transmitting optical system, a laser, and a laser driver, and the receiving module includes a receiving optical system, a detector, and an analog front end. During use, the laser in the transmitting module, under the control of the main control chip, emits a laser toward the target. When the laser hits the target, the target reflects the laser back to the receiving optical system and onto the detector. The analog front end then converts the analog signal into a digital signal. The main control chip then calculates and processes the signal to obtain a large number of discrete spatial coordinate points without topological structure, namely a point cloud. By processing the point cloud data through a computer, the vehicle can accurately perceive the current road conditions and make corresponding actions in a timely manner.

[0069] In an optional embodiment of the present application, the visual camera can be an ordinary camera or a binocular camera, and the present application does not impose any restrictions on this. Preferably, the binocular camera can obtain information such as the distance between vehicles from the depth of field image through a depth of field algorithm, which can be used to compare with the road perception data of the lidar, thereby improving data accuracy.

[0070] In the implementation manner of this application, sensors can be deployed throughout the entire road section, and when used, the entire road section can be divided into multiple areas or sub-sections. For example, visual cameras, lidars, and sound source sensors can be installed on road railings, which will not be elaborated in this application.

[0071] In addition, in the implementation manner of the present application, the types of collision accidents can generally include minor collisions and severe collisions. Minor collisions have little impact on traffic flow. Therefore, the existing traffic flow comparison method has large errors in the detection of minor collisions. In addition to traffic accidents, there are other factors that can cause similar changes, such as changes in traffic demand, temporary traffic control measures, etc. In the implementation manner of the present application, the fault type can be first judged based on the intensity of the sound generated by the sound source. That is, if the intensity of the collected sound is higher than the set threshold, the current collision accident type is determined to be a severe collision, otherwise it is a minor collision.

[0072] For example, since different sound source sensors are located at different positions, the sound signal intensities received for the same sound source are different. The sound data collected by all the deployed sound source sensors can be compared for the intensity of the sound generated by a certain sound source, and then it can be determined that the sound source sensor corresponding to the sound with the highest intensity is near the sound source. After determining the proximity of the sound source, it can be determined whether the sound with the highest intensity is higher than the set threshold. If it is higher, it is defined as a serious collision accident, and if it is lower, it is defined as a minor collision accident.

[0073] This application first defines the collision type based on sound, and then uses road perception data collected by multiple sensors to make a comprehensive judgment on the classification, avoiding the shortcomings of inaccurate detection and susceptibility to interference of traffic flow and video. The minor collisions and severe collisions in this application are explained separately below.

[0074] When a traffic collision is defined as a severe collision, the sound characteristics are more obvious. The sound data can be used in combination with the detection of the sound source sensor to identify the location of the sound source. Then, a comprehensive judgment can be made based on the sound source location and combined with other road perception data.

[0075] In this embodiment, the types of road perception data also include: vehicle motion data, the corresponding sensor includes at least one of a laser radar and a camera; Figure 2 As shown, step S3 specifically includes:

[0076] S31: If the current collision type is a severe collision, estimate the estimated azimuth angle and estimated distance of the current collision relative to the sensor based on the sound data;

[0077] S32: Determine accident information of the traffic collision accident based on the estimated azimuth angle and the estimated distance in combination with the vehicle motion data within a corresponding position range.

[0078] In an embodiment of the present application, the entire road section can be divided into multiple sub-sections, and multiple sound source sensors are set in each sub-section. The sound source sensors form a sensor array. Then, based on the relative delay estimation method, due to the geometric structure of the array, the signals received by each array have different degrees of delay. The relative delay estimation method estimates the delay difference between each array signal through cross-correlation, generalized cross-correlation (GCC) or phase difference, and then combines the geometric structure of the array to estimate the azimuth information of the sound source. This application does not elaborate on this.

[0079] In a preferred embodiment of the present application, the present application can set up a monitoring station in each sub-road section based on the above-mentioned TDOA algorithm principle, and at least one sound source sensor is arranged in the monitoring station. Then, a principle similar to the above is adopted, but because the sound source sensors are distributed throughout the road section, an array is not formed. At this time, since the sound intensity generated during a severe collision is relatively large, the sound source sensors at multiple monitoring stations can receive the sound signal generated during a severe collision. Therefore, based on the time that each sound source sensor corresponds to a received signal, the distance between each sound source sensor and the sound source can be calculated. By using a principle similar to the TDOA algorithm for calculation, the estimated azimuth and estimated distance between the current collision and the sensor can be obtained. Specifically, step S31 can be determined in the following manner:

[0080] The estimated azimuth angle and estimated distance of the current collision relative to the sensor are estimated based on the sound data, such as Figure 3 Shown, including:

[0081] S311: Based on a time delay estimation algorithm, determine the distance between the sound source and each sound source sensor with one of the sound source sensors as a reference origin;

[0082] S312: Calculate, based on the distance between the sound source and each sound source sensor, a first distance between the sound source and the i-th sound source sensor and a second distance between the sound source and the j-th sound source sensor, and determine a distance difference between the first distance and the second distance; i and j are different and both less than N;

[0083] S313: Determine an estimated azimuth angle and an estimated distance between the current collision and the sensor based on the speed of sound and each distance difference.

[0084] Assume that there are M randomly distributed base stations in a two-dimensional plane, receiving signals transmitted by the same target signal source. The base station position coordinates are (xi, yi), i = 1, 2, 3, ..., m, and the target position coordinates are (x, y). Assume that the received signal of the i-th base station is:

[0085] ui(t)=s(t-di)+vi(t) (1)

[0086] Where s(t) is the source signal transmitted by the target, di is the time delay of the original signal propagating to the i-th base station, vi(t) is the additive white Gaussian noise, and it is assumed that the signal and noise are independent of each other.

[0087] First, using the delay of the first base station as a reference, the generalized cross-correlation (GCC) algorithm is used to estimate the time difference between the source signal arriving at the second, third to M base stations and the first base station:

[0088] di1=di-d1, i=2, 3 to M(2)

[0089] Then, using the time difference d, according to the geometric relationship between the base station and the target, the equations are set to solve the target position. Specifically, according to the time difference estimation di1 and the speed of light c, the distance difference between the target and the i-th base station and the first base station can be obtained: ri1=cdi1=ri-r1, i=2,3 to M

[0090] Among them, ri is the distance from the target to the i-th base station. The coordinates and the above formula are used to construct a nonlinear equation system. Then the positioning problem becomes a problem of solving the equations, and the target position can be obtained using the least squares method.

[0091] This application does not elaborate on the TDOA algorithm. It can be seen that by combining the TDOA algorithm and treating each base station as a microphone in a "microphone array", the signal can be first located within the plane of the entire traffic section.

[0092] Furthermore, the vehicle motion data includes: speed, position, heading angle and distance between adjacent vehicles at each moment, and the accident information includes vehicle images, accident videos and point cloud data; correspondingly, the sensors include lidar and cameras, which are not elaborated here.

[0093] For serious collision accidents, the present application can determine the accident information of the traffic collision accident based on the estimated azimuth and estimated distance after determining the estimated azimuth and estimated distance through sound, combined with the vehicle motion data within the corresponding position range, specifically, Figure 4 Shown, including:

[0094] S321: Within a position range corresponding to the azimuth and distance of the current collision relative to the sensor, if a change in the speed, position, heading angle, and distance between adjacent vehicles within a set time period exceeds a corresponding threshold, determining that a traffic collision has occurred within the corresponding position range;

[0095] S322: Retrieve the point cloud data collected by the lidar within the corresponding position range, as well as the vehicle image and accident video of the relevant vehicle in the image captured by the camera.

[0096] It should be noted that the accident information of a traffic collision accident generally includes attribute information of the relevant vehicles involved in the collision, such as body color, vehicle license plate number, etc. Each vehicle can be anchored through the above attribute information, so that the detection data of sensors such as cameras and lidars can be further combined to calculate the vehicle distance, position, heading angle, speed and azimuth, etc., and due to serious collision accidents, the traffic volume will be different from that during normal traffic, but the difference in traffic volume is easily affected by other factors, such as sudden lightning strikes and rainy weather resulting in reduced visibility, which is similar to the parameters formed by the serious collision of this application, forming a certain induction. This application determines whether a collision has occurred by the change in the distance between vehicles in the area, cleverly taking advantage of the suddenness of traffic accidents and excluding factors such as weather and other continuous conditions, thereby improving the detection accuracy of collision accidents.

[0097] For minor accidents, on the one hand, the sound intensity generated is low, and on the other hand, the traffic volume changes are also small. At this time, the location range of the suspected traffic collision accident can be determined based on the distance between adjacent vehicles in the entire road section at each moment, combined with the speed, position, and heading angle of each vehicle; then, within the location range of the suspected traffic collision accident, based on the sound data in the location range, it is judged whether the suspected traffic collision accident was misjudged. If not, it is determined that a traffic collision accident occurred within the location range.

[0098] That is, when this application deals with a minor accident, it first determines the location range of the suspected traffic accident, and then "assumes" that an accident has occurred there, and compares the sound data here to see if it is the sound data when the accident occurred, thereby assisting in determining whether there is a fault.

[0099] Specifically, in a preferred embodiment of the present application, historical data can be combined for analysis. Specifically, the location range of the suspected traffic collision accident is determined based on the distance between adjacent vehicles in the entire road section at each moment, combined with the speed, position, and heading angle of each vehicle, such as Figure 5 Shown, including:

[0100] S3211: Split the road section to be detected into multiple sub-sections, and obtain historical vehicle distance datasets of the multiple sub-sections;

[0101] S3212: For each sub-segment, a correlation analysis is performed on the change in the distance between adjacent vehicles in the current time period within a set time period and the change in the distance between adjacent vehicles in the historical vehicle distance dataset for the sub-segment when there are no accidents and when there are accidents, to determine whether the change in the current sub-segment is abnormal.

[0102] S3213: If abnormal, determine whether the consistency of the speed change, position change, and heading angle change of each vehicle is higher than a set threshold based on the speed, position, and heading angle of each vehicle;

[0103] S3214: If yes, determine the corresponding sub-road segment as the location range of the suspected traffic collision accident.

[0104] For example, assuming that the road section is divided into four sub-sections: a, b, c and d, then for each sub-section, the changing trends of the historical data of the vehicle spacing are compared. Although the change in the vehicle spacing is small, the present application creatively compares the changing trends of the historical vehicle spacing with the current changing trends. If the consistency of the changing trends of at least one of the four sub-sections is higher than the set threshold, the corresponding sub-section is determined to be an accident section.

[0105] Furthermore, this application can also use correlation analysis of changing trends. Because road sections closer to the accident site are more affected by the accident and have lower consistency in changing trends, road sections farther from the accident site are less affected and have higher consistency in changing trends. Based on this, this application combines historical changing trends and uses the consistency of changing trends to accurately determine minor collisions.

[0106] For example, assuming that a minor collision occurs in sub-section b, the change trend consistency of section b is the lowest, followed by a and c, and the change trend consistency of section d is the highest. Therefore, combined with the changes in the consistency trend, it can be determined that a minor collision occurred in sub-section b.

[0107] It can be understood that this application coordinates the road perception data of multiple sensors, then first determines the collision type based on the sound, and then combines all road perception data to comprehensively judge whether it is a collision accident. It solves the shortcomings of traffic flow comparison and video detection, and provides road traffic management departments with richer accident information to facilitate timely processing.

[0108] Furthermore, the present application provides a traffic collision accident detection device, such as Figure 6 Shown, including:

[0109] Acquisition module 1, acquires multiple types of road perception data of the road section to be detected, wherein the types of the road perception data include sound, and each type of road perception data is collected by a corresponding sensor;

[0110] Type determination module 2, determining the type of the current collision accident based on the intensity of the sound;

[0111] The accident information determination module 3 determines the accident information of the traffic collision accident according to the current collision accident type and in combination with the multiple road perception data.

[0112] It can be understood that the traffic collision accident detection device provided by this application coordinates the road perception data of multiple sensors, and then first determines the collision type based on the sound, and then combines all road perception data to comprehensively judge whether it is a collision accident. It solves the drawbacks of traffic flow comparison and video detection, and provides road traffic management departments with richer accident information to facilitate timely processing.

[0113] From a hardware perspective, the present invention provides an embodiment of an electronic device for implementing all or part of the traffic collision accident detection method. The electronic device specifically includes the following:

[0114] A processor, a memory, a communications interface, and a bus; wherein the processor, memory, and communications interface communicate with each other via the bus; the communications interface is used to implement information transmission between related devices such as servers, devices, distributed message middleware cluster devices, various databases, and user terminals; the electronic device can be a desktop computer, a tablet computer, a mobile terminal, etc., but this embodiment is not limited thereto. In this embodiment, the electronic device can be implemented with reference to the embodiments of the traffic collision accident detection method and the embodiments of the traffic collision accident detection device in the embodiments, the contents of which are incorporated herein and repeated parts are not repeated.

[0115] Figure 7 FIG. 9 is a schematic block diagram of the system structure of the electronic device 9600 according to an embodiment of the present invention. Figure 7 As shown, the electronic device 9600 may include a central processing unit 9100 and a memory 9140; the memory 9140 is coupled to the central processing unit 9100. It is worth noting that the Figure 7 is exemplary; other types of structures may also be used to supplement or replace this structure to implement telecommunication functions or other functions.

[0116] In one embodiment, the traffic collision accident detection function can be integrated into the central processing unit 9100.

[0117] In another embodiment, the traffic collision accident detection device can be configured separately from the central processing unit 9100. For example, the traffic collision accident detection device can be configured as a chip connected to the central processing unit 9100, and the traffic collision accident detection function can be realized through the control of the central processing unit.

[0118] like Figure 7 As shown, the electronic device 9600 may further include: a communication module 9110, an input unit 9120, an audio processor 9130, a display 9160, and a power supply 9170. It is worth noting that the electronic device 9600 does not necessarily have to include Figure 7 In addition, the electronic device 9600 may also include all components shown in Figure 7 For components not shown, reference may be made to the prior art.

[0119] like Figure 7 As shown, the central processing unit 9100 is sometimes also referred to as a controller or operation control, and may include a microprocessor or other processor device and / or logic device. The central processing unit 9100 receives input and controls the operation of various components of the electronic device 9600.

[0120] Memory 9140 can be, for example, one or more of a cache, flash memory, hard drive, removable media, volatile memory, non-volatile memory, or other suitable devices. It can store the aforementioned failure-related information and also store programs that execute the relevant information. The CPU 9100 can execute the programs stored in memory 9140 to implement information storage or processing.

[0121] The input unit 9120 provides input to the central processing unit 9100. The input unit 9120 may be, for example, a keypad or touch input device. The power supply 9170 is used to provide power to the electronic device 9600. The display 9160 is used to display objects such as images and text. The display may be, for example, an LCD display, but is not limited thereto.

[0122] The memory 9140 may be a solid-state memory, such as a read-only memory (ROM), a random access memory (RAM), or a SIM card. Alternatively, it may be a memory that retains information even when power is off, can be selectively erased, and is provided with more data. Examples of such memory are sometimes referred to as EPROMs. The memory 9140 may also be some other type of device. The memory 9140 includes a buffer memory 9141 (sometimes referred to as a buffer). The memory 9140 may include an application / function storage unit 9142 for storing application programs and function programs or processes for executing the operation of the electronic device 9600 by the central processing unit 9100.

[0123] The memory 9140 may also include a data storage unit 9143 for storing data, such as contacts, digital data, pictures, sounds, and / or any other data used by the electronic device. The driver storage unit 9144 of the memory 9140 may include various driver programs for communication functions of the electronic device and / or for executing other functions of the electronic device (such as messaging applications, address book applications, etc.).

[0124] The communication module 9110 is a transmitter / receiver 9110 that transmits and receives signals via an antenna 9111. The communication module (transmitter / receiver) 9110 is coupled to the central processor 9100 to provide input signals and receive output signals, which may be the same as in a conventional mobile communication terminal.

[0125] Based on different communication technologies, multiple communication modules 9110 can be provided in the same electronic device, such as a cellular network module, a Bluetooth module, and / or a wireless local area network module. The communication module (transmitter / receiver) 9110 is also coupled to a speaker 9131 and a sound source sensor 9132 via an audio processor 9130 to provide audio output via the speaker 9131 and receive audio input from the sound source sensor 9132, thereby implementing common telecommunication functions. The audio processor 9130 may include any suitable buffer, decoder, amplifier, etc. In addition, the audio processor 9130 is also coupled to the central processing unit 9100, enabling local recording via the sound source sensor 9132 and playback of stored sounds via the speaker 9131.

[0126] An embodiment of the present invention also provides a computer-readable storage medium capable of implementing all steps of the traffic collision accident detection method in the above embodiment, where the execution entity can be a server. The computer-readable storage medium stores a computer program, which, when executed by a processor, implements all steps of the traffic collision accident detection method in the above embodiment.

[0127] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, apparatus, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0128] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (apparatus), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as a combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0129] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0130] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0131] Specific embodiments are used in the present invention to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core ideas. At the same time, for those skilled in the art, according to the ideas of the present invention, there may be changes in the specific implementation methods and application scopes. In summary, the contents of this specification should not be understood as limiting the present invention.

Claims

1. A traffic collision accident detection method, characterized in that: include: Acquire multiple types of road perception data for the road section to be detected, where the types of the road perception data include sound, and each type of road perception data is collected by a corresponding sensor; determining the type of the current collision accident according to the intensity of the sound; Determining accident information of a traffic collision accident based on the current collision accident type and combining the multiple road perception data; The collision accident types include minor collisions and severe collisions. Correspondingly, determining the current collision accident type based on the intensity of the sound includes: if the intensity of the collected sound is higher than a set threshold, determining the current collision accident type as a severe collision; otherwise, determining the current collision accident type as a minor collision; The types of road perception data also include: vehicle motion data, corresponding sensors include lidar and camera; the accident information of the traffic collision accident is determined based on the current collision accident type and combined with the multiple road perception data, including: If the current collision accident type is a severe collision, estimating an estimated azimuth angle and an estimated distance of the current collision relative to the sensor based on the sound data; determining accident information of the traffic collision accident based on the estimated azimuth angle and the estimated distance in combination with the vehicle motion data within the corresponding position range; If the current collision accident type is a minor collision, a suspected traffic collision accident is determined based on all vehicle motion data in the entire road section; and whether a traffic collision accident has occurred is determined based on the sound data within the location range of each suspected traffic collision accident; If a traffic collision is determined to have occurred, the point cloud data collected by the lidar within the corresponding position range, as well as the vehicle images of the relevant vehicles in the camera images and the accident video are retrieved.

2. The traffic collision accident detection method according to claim 1, characterized in that: The sensor includes a plurality of sound source sensors distributed at a plurality of monitoring sites, and the estimating, based on the sound data, an estimated azimuth angle and an estimated distance of the current collision relative to the sensor includes: Based on the time delay estimation algorithm, the distance between the sound source and each sound source sensor is determined with one of the sound source sensors as the reference origin; Calculating, based on the distance between the sound source and each sound source sensor, a first distance between the sound source and the i-th sound source sensor and a second distance between the sound source and the j-th sound source sensor, and determining a distance difference between the first distance and the second distance; i and j are different and both less than N; An estimated azimuth and an estimated distance of the current collision relative to the sensor are determined based on the speed of sound and each distance difference.

3. The traffic collision accident detection method according to claim 1, characterized in that: The vehicle motion data includes: speed, position, heading angle and distance between adjacent vehicles at each moment; the accident information includes vehicle images, accident videos and point cloud data; The determining of the accident information of the traffic collision accident based on the estimated azimuth and the estimated distance in combination with the vehicle motion data within the corresponding position range includes: If, within a position range corresponding to the azimuth and distance of the current collision relative to the sensor, the change in the speed, position, heading angle, and distance between adjacent vehicles within a set time period exceeds a corresponding threshold, it is determined that a traffic collision accident has occurred within the corresponding position range; Retrieve the point cloud data collected by the lidar within the corresponding position range, as well as the vehicle image and accident video of the relevant vehicle in the camera image.

4. The traffic collision accident detection method according to claim 1, characterized in that: The vehicle motion data includes: speed, position, heading angle, and the distance between adjacent vehicles; the determination of whether a traffic collision has occurred based on the sound data within the location range of each suspected traffic collision accident includes: Determine the location range of a suspected traffic collision based on the distance between adjacent vehicles in the entire road section at each moment, combined with the speed, position, and heading angle of each vehicle; Within the location range of the suspected traffic collision accident, it is determined whether the suspected traffic collision accident is a misjudgment based on the sound data in the location range. If not, it is determined that a traffic collision accident has occurred within the location range.

5. The traffic collision accident detection method according to claim 4, characterized in that: Determining the location range of a suspected traffic collision accident based on the distance between adjacent vehicles in the entire road section at each moment, combined with the speed, position, and heading angle of each vehicle, includes: Splitting the road section to be detected into multiple sub-sections, and obtaining a historical vehicle distance dataset of the multiple sub-sections; For each sub-segment, the change in the distance between adjacent vehicles in the current period within a set time period is analyzed for correlation with the change in the historical vehicle distance data set for the sub-segment when there are no accidents and when there are accidents, to determine whether the change in the current sub-segment is abnormal; If there is an abnormality, the speed, position and heading angle of each vehicle are combined to determine whether the consistency of the speed change, position change and heading angle change of each vehicle is higher than the set threshold; If so, the corresponding sub-segment is determined as the location range of the suspected traffic collision accident.

6. A traffic collision accident detection device, wherein the traffic collision accident detection device implements the method according to any one of claims 1 to 5, characterized in that: The traffic collision accident detection device comprises: an acquisition module for acquiring a variety of road perception data of the road section to be detected, wherein the types of the road perception data include sound, and each type of road perception data is collected by a corresponding sensor; A type determination module, which determines the type of the current collision accident based on the intensity of the sound; The accident information determination module determines the accident information of the traffic collision accident based on the current collision accident type and in combination with the multiple road perception data.

7. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the method according to any one of claims 1 to 5 is implemented.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

Images