Vehicle overspeed automatic detection method and system based on radar and video integrated machine

Abstract

The application discloses a vehicle overspeed automatic detection method and system based on a radar and video integrated machine. According to radar detection data, time stamp information and vehicle distance information of a target vehicle corresponding to each frame are obtained. According to time stamp information and vehicle distance information of the target vehicle corresponding to adjacent frames, a plurality of vehicle speeds are obtained. The plurality of vehicle speeds are sorted to obtain a plurality of preprocessed vehicle speeds, and the plurality of preprocessed vehicle speeds are subjected to mean value processing to obtain a target speed of the target vehicle. If the target speed is greater than a threshold speed, the target vehicle is an overspeed vehicle. According to the radar detection data, vehicle position center point coordinates of the overspeed vehicle are obtained. The vehicle position center point coordinates are subjected to coordinate system matrix conversion to obtain vehicle position center point coordinates in an image coordinate system. According to video image data, an overspeed vehicle image centered on the vehicle position center point coordinates in the image coordinate system is obtained, vehicle identification is performed on the overspeed vehicle image, and vehicle information of the overspeed vehicle is obtained.

Description

Technical Field

[0001] This application relates to the field of intelligent transportation technology, and in particular to an automatic vehicle overspeed detection method and system based on a radar-visual integrated machine. Background Technology

[0002] Traffic safety is a crucial factor affecting people's lives. Intersections, as the most congested points of vehicle and pedestrian traffic, make traffic safety at intersections paramount. Whether vehicles comply with traffic regulations and do not exceed speed limits when crossing intersections is a vital factor in ensuring traffic safety at intersections.

[0003] Traditional methods use induction coils for speed measurement. They detect vehicles by observing the change in inductance of the coil when a vehicle passes through a magnetic field, and calculate the vehicle's speed based on the time difference. However, induction coils in traditional methods are easily damaged by heavy vehicles, road surface cracks, deformation, and during repairs. They are also susceptible to external factors, resulting in low detection accuracy and a high risk of misjudgment.

[0004] Application content

[0005] The purpose of this application is to solve the technical problem of low detection accuracy caused by traditional methods that use changes in the inductance of induction coils for detection. To achieve the above objective, this application provides an automatic vehicle overspeed detection method and system based on a radar-visual integrated machine.

[0006] This application provides an automatic vehicle overspeed detection method based on a radar-visual integrated machine, including:

[0007] Acquire radar detection data and video image data;

[0008] Based on the radar detection data, obtain the timestamp information and vehicle distance information of the target vehicle for each frame;

[0009] Multiple vehicle speeds are obtained based on the timestamp information and distance information of the target vehicle corresponding to adjacent frames;

[0010] The speeds of the multiple vehicles are sorted to obtain multiple preprocessed vehicle speeds, and the average of the multiple preprocessed vehicle speeds is applied to obtain the target speed of the target vehicle; wherein, the speeds of the multiple preprocessed vehicle speeds are greater than the speeds of all other vehicles among the multiple vehicle speeds.

[0011] If the target speed is greater than the threshold speed, the target vehicle is a speeding vehicle. Based on the radar detection data, the coordinates of the center point of the vehicle position in the radar coordinate system are obtained.

[0012] Perform coordinate system matrix transformation on the vehicle position center point coordinates in the radar coordinate system to obtain the vehicle position center point coordinates in the image coordinate system.

[0013] Based on the video image data, an image of a speeding vehicle centered on the vehicle's position center point in the image coordinate system is obtained. Vehicle identification is performed on the speeding vehicle image to obtain the vehicle information of the speeding vehicle. The speeding vehicle image corresponds to the same timestamp as the vehicle's position center point coordinate in the radar coordinate system.

[0014] In one embodiment, performing coordinate system matrix transformation on the vehicle position center point coordinates to obtain the vehicle position center point coordinates in the image coordinate system includes:

[0015] The coordinates of the vehicle's center point are transformed using a camera coordinate system matrix to obtain the coordinates of the vehicle's center point in the camera coordinate system.

[0016] The coordinates of the vehicle's center point in the camera coordinate system are transformed by an image coordinate system matrix to obtain the coordinates of the vehicle's center point in the image coordinate system.

[0017] In one embodiment, the step of acquiring an image of a speeding vehicle centered on the vehicle's center point in the image coordinate system based on the video image data, and performing vehicle recognition on the speeding vehicle image to obtain vehicle information of the speeding vehicle includes:

[0018] Based on the video image data, obtain an image of the speeding vehicle with a length of 300 to 500 pixels and a width of 200 to 400 pixels, centered on the vehicle's position center point in the image coordinate system.

[0019] Based on the target detection algorithm, the target region of the speeding vehicle image is extracted to obtain the license plate image;

[0020] The license plate image is identified using a license plate recognition algorithm to obtain the license plate information of the speeding vehicle.

[0021] In one embodiment, obtaining the timestamp information and vehicle distance information of the target vehicle corresponding to each frame based on the radar detection data includes:

[0022] According to the Kalman filter tracking algorithm, the radar detection data is used to track the target vehicle, and the timestamp information and distance information of the target vehicle corresponding to each frame are obtained.

[0023] In one embodiment, the automatic vehicle overspeed detection method based on a radar-visual integrated machine further includes:

[0024] Traffic management is performed on the speeding vehicle based on the timestamp information, distance information, target speed, image of the speeding vehicle, and vehicle information.

[0025] In one embodiment, this application provides a vehicle overspeed automatic detection system based on a radar-visual integrated machine, comprising:

[0026] The data acquisition module is used to acquire the radar detection data and the video image data;

[0027] The target vehicle information processing module is used to obtain the timestamp information and vehicle distance information of the target vehicle corresponding to each frame based on the radar detection data.

[0028] The vehicle speed processing module is used to obtain multiple vehicle speeds based on the timestamp information and vehicle distance information of the target vehicle corresponding to adjacent frames;

[0029] The target speed processing module is used to sort the multiple vehicle speeds to obtain multiple pre-processed vehicle speeds, and to perform average processing on the multiple pre-processed vehicle speeds to obtain the target speed of the target vehicle; wherein, the multiple pre-processed vehicle speeds are greater than the speeds of all other vehicles among the multiple vehicle speeds.

[0030] The overspeed vehicle coordinate position processing module is used to determine if the target vehicle is an overspeed vehicle if the target speed is greater than a threshold speed, and to obtain the coordinates of the center point of the overspeed vehicle's position based on the radar detection data.

[0031] The coordinate transformation module is used to perform coordinate system matrix transformation on the coordinates of the vehicle's center point to obtain the coordinates of the vehicle's center point in the image coordinate system.

[0032] The speeding vehicle information acquisition module is used to acquire a speeding vehicle image centered on the vehicle's position center point coordinates in the image coordinate system based on the video image data, perform vehicle recognition on the speeding vehicle image, and obtain the vehicle information of the speeding vehicle.

[0033] In one embodiment, the coordinate transformation module includes:

[0034] The camera coordinate system transformation module is used to perform a camera coordinate system matrix transformation on the coordinates of the vehicle's center point to obtain the coordinates of the vehicle's center point in the camera coordinate system.

[0035] The image coordinate system transformation module is used to perform an image coordinate system matrix transformation on the coordinates of the vehicle's center point in the camera coordinate system to obtain the coordinates of the vehicle's center point in the image coordinate system.

[0036] In one embodiment, the speeding vehicle information acquisition module includes:

[0037] The speeding vehicle image processing module is used to acquire, based on the video image data, an image of the speeding vehicle with a length of 300 to 500 pixels and a width of 200 to 400 pixels, centered on the vehicle's position center point in the image coordinate system.

[0038] The license plate image processing module is used to extract the target region from the speeding vehicle image according to the target detection algorithm to obtain the license plate image;

[0039] The license plate information processing module is used to identify the license plate image according to the license plate recognition algorithm to obtain the license plate information of the speeding vehicle.

[0040] In one embodiment, the target vehicle information processing module includes:

[0041] The timestamp vehicle distance acquisition module is used to track target vehicles based on the radar detection data according to the Kalman filter tracking algorithm, and obtain the timestamp information and vehicle distance information of the target vehicle corresponding to each frame.

[0042] In one embodiment, the vehicle overspeed automatic detection system based on the radar-visual integrated machine further includes:

[0043] The radar-visual integrated machine is set up at an intersection to collect radar detection data and video image data and transmit them to the data acquisition module;

[0044] A data storage server is used to store the timestamp information, vehicle distance information, target speed, image of the speeding vehicle, and vehicle information of the speeding vehicle.

[0045] The management platform is used to manage traffic for the speeding vehicle based on the timestamp information, distance information, target speed, image of the speeding vehicle, and vehicle information.

[0046] The aforementioned automatic vehicle speeding detection method and system based on a radar-visual integrated machine acquires radar detection data and video image data. Based on the radar detection data, it tracks vehicles to determine if they are speeding. If a vehicle is speeding, it is identified as a speeding vehicle, and the corresponding video image data is retrieved. The radar-tracked data is mapped onto the video image to accurately locate the corresponding speeding vehicle image information. Vehicle information is then extracted from the speeding vehicle image to accurately obtain vehicle information about the speeding vehicle. Therefore, based on the radar detection data, video image data, and analyzed speeding vehicle information, traffic management is performed on speeding vehicles, accurately achieving automatic speeding vehicle detection, improving the accuracy of automatic speeding vehicle detection, and providing accurate data for traffic management systems.

[0047] Therefore, the automatic vehicle speeding detection method and system based on radar-video integrated machine provided in this application can achieve higher accuracy in judging vehicle speeding and higher accuracy in identifying information related to speeding vehicles. By combining the advantages of radar sensors and video sensors, the automatic vehicle speeding detection method and system based on radar-video integrated machine provided in this application improves the accuracy of speed measurement data and captured license plate data, achieving the goal of both accurate speed measurement and accurate vehicle information. This improves the efficiency of information integration and provides a more reliable basis for traffic management. Attached Figure Description

[0048] Figure 1 This is a flowchart illustrating the steps of the automatic vehicle speeding detection method based on the radar-visual integrated machine provided in this application.

[0049] Figure 2 This is a structural schematic diagram of the vehicle overspeed automatic detection system based on the radar-visual integrated machine provided in this application. Detailed Implementation

[0050] The technical solution of this application will be further described in detail below with reference to the accompanying drawings and embodiments.

[0051] Please see Figure 1 This application provides an automatic vehicle overspeed detection method based on a radar-visual integrated machine, comprising:

[0052] S10, acquire radar detection data and video image data;

[0053] S20: Based on radar detection data, obtain the timestamp information and vehicle distance information of the target vehicle for each frame;

[0054] S30: Based on the timestamp information and distance information of the target vehicle corresponding to adjacent frames, obtain multiple vehicle speeds;

[0055] S40: Sort the speeds of multiple vehicles to obtain multiple preprocessed vehicle speeds, and average the multiple preprocessed vehicle speeds to obtain the target speed of the target vehicle; wherein, the speeds of the multiple preprocessed vehicles are greater than the speeds of the vehicles other than the speeds of the multiple vehicles.

[0056] S50, if the target speed is greater than the threshold speed, the target vehicle is a speeding vehicle. Based on the radar detection data, the coordinates of the center point of the speeding vehicle in the radar coordinate system are obtained.

[0057] S60, perform coordinate system matrix transformation on the coordinates of the vehicle's center point to obtain the coordinates of the vehicle's center point in the image coordinate system;

[0058] S70: Based on video image data, acquire an image of a speeding vehicle centered on the vehicle's position center point in the image coordinate system, perform vehicle identification on the speeding vehicle image, and obtain vehicle information of the speeding vehicle; wherein, the speeding vehicle image and the vehicle's position center point coordinates in the radar coordinate system correspond to the same timestamp information.

[0059] In this embodiment, the radar-video integrated unit is used to perceive and collect physical information data of vehicles at an intersection. Radar detection data can include radar-based vehicle information such as vehicle timestamps, vehicle distance, vehicle orientation, vehicle speed, vehicle acceleration, and vehicle spatial position. Based on video image data, vehicle attribute information (such as license plate number, license plate color, and vehicle color) and the time information of each frame can be obtained. Based on the radar detection data, the timestamp information and vehicle distance information corresponding to the target vehicle in each frame detected by the radar are extracted. The radar-video integrated unit's parameters are configured via a host computer, setting the system time and the world coordinates of the radar-video integrated unit, and setting an overspeed threshold. The radar detects the movement information of all vehicles at the intersection in real time, and the camera collects vehicle image information at the intersection in real time.

[0060] The timestamps for the target vehicle in each frame are t1, t2, t3...t n The corresponding vehicle distance information is s1, s2, s3...s n Based on the timestamp information and distance information of the target vehicle between two adjacent frames, the vehicle speed is obtained. For example, the vehicle speed information v1 extracted from the first and second frames is (s2-s1) / (t2-t1), and so on, to obtain the vehicle speed information extracted from the (n+1)th frame and the nth frame, thus obtaining multiple vehicle speed information v1, v2, v3...v n .

[0061] Based on the obtained vehicle speed information v1, v2...v nThe vehicle speeds are sorted to obtain the three, four, or five vehicle speeds with the largest values, forming multiple pre-processed vehicle speeds. These pre-processed vehicle speeds are then averaged to obtain the target speed of the target vehicle. For example, in one embodiment, the three largest speed values ​​v1, v2, and v3 are averaged to obtain the final target speed v = (v1 + v2 + v3) / 3.

[0062] The target vehicle's speed is compared to an overspeed threshold speed. If the target speed exceeds the threshold speed, the target vehicle is considered speeding. Therefore, the timestamp information, distance information, and speed of the speeding vehicle are stored based on radar data to provide accurate and reliable data for subsequent law enforcement platforms. The center coordinates P(R, a) of the speeding vehicle are obtained by averaging the positions of multiple coordinate points representing the vehicle's location. P(R, a) represents the position coordinates of the speeding vehicle in the radar coordinate system, where R represents the radius and a represents the angle.

[0063] Based on the coordinate transformation matrix from the radar coordinate system to the camera coordinate system, the coordinates P1(X, Y, Z) of the vehicle's center point in the camera coordinate system are obtained. The coordinate transformation matrix from the radar coordinate system to the camera coordinate system is:

[0064]

[0065] Based on the coordinate transformation matrix from the camera coordinate system to the image coordinate system, the coordinates of the vehicle's center point P2(X, Y) in the image coordinate system are obtained. f represents the focal length. The coordinate transformation matrix from the camera coordinate system to the image coordinate system is:

[0066]

[0067] Based on the acquired video image data, images corresponding to the same timestamp as the calculated center point coordinates of the speeding vehicle are extracted. The video image data and radar detection data are acquired synchronously in time, corresponding to the same timestamp information. The camera and radar have the same sampling rate, ensuring temporal synchronization between the video image data and the radar detection data. This ensures that the center point coordinates of the speeding vehicle and the image of the speeding vehicle centered on the center point coordinates in the image coordinate system both correspond to the same timestamp. Images with the same timestamp information are extracted from the video image data, and images of the speeding vehicle centered on the center point coordinates in the image coordinate system are obtained. The image of the speeding vehicle centered on the center point coordinates in the image coordinate system can be understood as expanding outwards from the center point coordinates in the image coordinate system, extracting the region of interest corresponding to a certain range, and forming the speeding vehicle image. The speeding vehicle image can contain the overall outline of the speeding vehicle. Therefore, by detecting and identifying the speeding vehicle in the image, relevant vehicle attribute information can be obtained, such as vehicle model, vehicle color, and vehicle license plate number.

[0068] The automatic vehicle speeding detection method based on a radar-visual integrated machine provided in this application acquires radar detection data and video image data through the integrated machine, tracks vehicles based on the radar detection data, and determines whether a vehicle is speeding. If a vehicle is speeding, it is identified as a speeding vehicle, and the corresponding video image data is retrieved. The radar-tracked data is mapped onto the video image to accurately locate the corresponding speeding vehicle image information, and vehicle information is extracted from the speeding vehicle image to accurately obtain vehicle information about the speeding vehicle. Therefore, based on the radar detection data, video image data, and the analyzed speeding vehicle information data corresponding to speeding vehicles, traffic management is carried out for speeding vehicles, accurately realizing automatic detection of speeding vehicles, improving the accuracy of automatic speeding vehicle detection, and providing accurate basis for traffic management systems.

[0069] Therefore, the automatic vehicle speeding detection method based on a radar-video integrated machine provided in this application achieves higher accuracy in judging vehicle speeding and higher accuracy in identifying information related to speeding vehicles. By combining the advantages of radar sensors and video sensors, this method improves the accuracy of speed measurement data and captured license plate data, achieving both accurate speed measurement and accurate vehicle information. This enhances the efficiency of information integration and provides a more reliable basis for traffic management.

[0070] In one embodiment, S20, based on radar detection data, the timestamp information and vehicle distance information of the target vehicle corresponding to each frame are obtained, including:

[0071] S210, based on the Kalman filter tracking algorithm, performs target vehicle tracking on radar detection data, and obtains the timestamp information and vehicle distance information of the target vehicle for each frame.

[0072] In this embodiment, the Kalman filter tracking algorithm is used to track all vehicle motion information detected by the radar, which improves the accuracy of radar detection data, resulting in high accuracy of vehicle speeding detection and accurate acquisition of the timestamp information and distance information of the target vehicle for each frame.

[0073] In one embodiment, S60, a coordinate system matrix transformation is performed on the coordinates of the vehicle's center point to obtain the coordinates of the vehicle's center point in the image coordinate system, including:

[0074] S610, perform camera coordinate system matrix transformation on the coordinates of the vehicle's center point to obtain the coordinates of the vehicle's center point in the camera coordinate system;

[0075] S620 performs an image coordinate system matrix transformation on the coordinates of the vehicle's center point in the camera coordinate system to obtain the coordinates of the vehicle's center point in the image coordinate system.

[0076] In this embodiment, the coordinates of the vehicle's center point in the radar coordinate system are transformed into the vehicle's center point in the image coordinate system through camera coordinate system matrix transformation and image coordinate system matrix transformation. This realizes the mapping transformation from radar detection data to video image data and makes full use of the mutual fusion of radar detection data and video image data.

[0077] In one embodiment, S70, based on video image data, an image of the speeding vehicle centered on the vehicle's position center point in the image coordinate system is obtained; vehicle recognition is performed on the speeding vehicle image to obtain vehicle information of the speeding vehicle, including:

[0078] S710, based on video image data, acquires an image of a speeding vehicle with a length of 300 to 500 pixels and a width of 200 to 400 pixels, centered on the vehicle's position center point in the image coordinate system.

[0079] S720 uses a target detection algorithm to extract the target region from images of speeding vehicles and obtain license plate images.

[0080] The S730 uses a license plate recognition algorithm to identify license plate images and obtain license plate information of speeding vehicles.

[0081] In this embodiment, images of speeding vehicles within a certain range are acquired using the vehicle's center point coordinates in the image coordinate system. The speeding vehicle images contain the speeding vehicles, and their specific length and width can be limited according to actual conditions. The target detection algorithm can include the YOLOv3 target detection model. Using a trained YOLOv3 target detection model, a target region can be extracted from a region of interest in the speeding vehicle images, obtaining a region of interest image. The target region can be a license plate area, the front of the vehicle, or the rear of the vehicle, etc., and the region of interest can be extracted accordingly.

[0082] License plate recognition algorithms can include CNN license plate recognition models. A trained CNN license plate recognition model can further detect and identify regions of interest (such as license plate images) to obtain the corresponding license plate number, thereby identifying which vehicle was speeding.

[0083] In one embodiment, by using object detection algorithms and license plate recognition algorithms, vehicle information can be extracted from each region of interest in an image of a speeding vehicle, such as vehicle color, vehicle model, or vehicle license plate number.

[0084] In this embodiment, the YOLOv3 object detection algorithm is used to accurately extract the license plate image of the speeding vehicle. The CNN license plate detection algorithm is then used to accurately extract the license plate number. Traffic management of the speeding vehicle is then carried out based on the license plate number of the speeding vehicle, and the law is enforced.

[0085] In one embodiment, the automatic vehicle speeding detection method based on the radar-visual integrated machine further includes:

[0086] S80 manages traffic for speeding vehicles based on timestamp information, distance information, target speed, images of speeding vehicles, and vehicle information.

[0087] In this embodiment, radar detection data and video image data, such as timestamp information, vehicle distance data, original images of speeding vehicles, speed of speeding vehicles in the region of interest, and license plate information of speeding vehicles, are obtained based on the methods and steps in the above embodiments to form a set of speeding evidence. This set of evidence is then uploaded to a data storage server and transmitted to the law enforcement platform to provide traffic police with evidence of speeding to punish speeding vehicles.

[0088] Please see Figure 2In one embodiment, this application provides an automatic vehicle speeding detection system 100 based on a radar-view integrated machine. The automatic vehicle speeding detection system 100 based on the radar-view integrated machine includes a data acquisition module 10, a target vehicle information processing module 20, a vehicle speed processing module 30, a target speed processing module 40, a speeding vehicle coordinate position processing module 50, a coordinate transformation module 60, and a speeding vehicle information acquisition module 70. The radar-view integrated machine is installed at an intersection to collect radar detection data and video image data, and transmits them to the data acquisition module 10. The data acquisition module 10 is used to acquire radar detection data and video image data. The target vehicle information processing module 20 is used to obtain the timestamp information and vehicle distance information of the target vehicle corresponding to each frame based on the radar detection data. The vehicle speed processing module 30 is used to obtain multiple vehicle speeds based on the timestamp information and vehicle distance information of the target vehicles corresponding to adjacent frames. The target speed processing module 40 is used to sort multiple vehicle speeds to obtain multiple pre-processed vehicle speeds, and to perform average processing on the multiple pre-processed vehicle speeds to obtain the target speed of the target vehicle; wherein, the multiple pre-processed vehicle speeds are greater than the speeds of the vehicles other than the multiple pre-processed vehicle speeds.

[0089] The speeding vehicle coordinate position processing module 50 is used to determine if the target vehicle is speeding if its speed exceeds a threshold speed, and to obtain the coordinates of the vehicle's center point based on radar detection data. The coordinate transformation module 60 performs a coordinate system matrix transformation on the vehicle's center point coordinates to obtain the vehicle's center point coordinates in the image coordinate system. The speeding vehicle information acquisition module 70 acquires an image of the speeding vehicle centered on the vehicle's center point coordinates in the image coordinate system based on video image data, performs vehicle recognition on the speeding vehicle image, and obtains the vehicle information of the speeding vehicle.

[0090] In this embodiment, the automatic vehicle speeding detection system 100 based on the radar-video integrated machine includes a radar video integrated machine, an algorithm server, a data storage server, and an enforcement platform connected together, with the radar facing the intersection. The algorithm server provides algorithm services, processes the vehicle physical information extracted from the radar video, and executes algorithms such as Kalman filtering to track target vehicles, YOLOv3 to detect vehicles, and CNN to recognize target vehicle license plate information. The algorithm server includes a data acquisition module 10, a target vehicle information processing module 20, a vehicle speed processing module 30, a target speed processing module 40, a speeding vehicle coordinate position processing module 50, a coordinate transformation module 60, and a speeding vehicle information acquisition module 70.

[0091] The description of the data acquisition module 10 can be found in the description of S10 in the above embodiments. The description of the target vehicle information processing module 20 can be found in the description of S20 in the above embodiments. The description of the vehicle speed processing module 30 can be found in the description of S30 in the above embodiments. The description of the target speed processing module 40 can be found in the description of S40 in the above embodiments. The description of the speeding vehicle coordinate position processing module 50 can be found in the description of S50 in the above embodiments. The description of the coordinate transformation module 60 can be found in the description of S60 in the above embodiments. The description of the speeding vehicle information acquisition module 70 can be found in the description of S70 in the above embodiments.

[0092] In one embodiment, the coordinate transformation module 60 includes a camera coordinate system transformation module and an image coordinate system transformation module. The camera coordinate system transformation module performs a camera coordinate system matrix transformation on the coordinates of the vehicle's center point to obtain the vehicle's center point coordinates in the camera coordinate system. The image coordinate system transformation module performs an image coordinate system matrix transformation on the vehicle's center point coordinates in the camera coordinate system to obtain the vehicle's center point coordinates in the image coordinate system.

[0093] In this embodiment, the description of the camera coordinate system transformation module can be found in the description of S610 in the above embodiment. The description of the image coordinate system transformation module can be found in the description of S620 in the above embodiment.

[0094] In one embodiment, the speeding vehicle information acquisition module 70 includes a speeding vehicle image processing module, a license plate image processing module, and a license plate information processing module. The speeding vehicle image processing module acquires a speeding vehicle image, centered on the vehicle's center point in the image coordinate system, with a length of 300 to 500 pixels and a width of 200 to 400 pixels, based on video image data. The license plate image processing module extracts the target region from the speeding vehicle image using a target detection algorithm to obtain the license plate image. The license plate information processing module identifies the license plate image using a license plate recognition algorithm to obtain the license plate information of the speeding vehicle.

[0095] In this embodiment, the description of the speeding vehicle image processing module can be found in the description of S710 in the above embodiment. The description of the license plate image processing module can be found in the description of S720 in the above embodiment. The description of the license plate information processing module can be found in the description of S730 in the above embodiment.

[0096] In one embodiment, the target vehicle information processing module 20 includes a timestamp vehicle distance acquisition module. The timestamp vehicle distance acquisition module is used to track target vehicles based on radar detection data using a Kalman filter tracking algorithm, obtaining the timestamp information and vehicle distance information of the target vehicle for each frame.

[0097] In this embodiment, the description of the timestamp vehicle distance acquisition module can be found in the description of S210 in the above embodiment.

[0098] In one embodiment, the automatic vehicle speeding detection system 100 based on the radar-visual integrated machine further includes a data storage server and a management platform. The data storage server is used to store timestamp information, vehicle distance information, target speed, images of speeding vehicles, and vehicle information. The management platform is used to perform traffic management for speeding vehicles based on the timestamp information, vehicle distance information, target speed, images of speeding vehicles, and vehicle information.

[0099] In this embodiment, the descriptions of the data storage server and management platform can be found in the description of S80 in the previous embodiment. The data storage server is used to store evidence of speeding vehicles, such as vehicle distance information, time information, speed information of speeding vehicles, image information of speeding vehicles, license plate information of speeding vehicles, etc. The enforcement platform is used by traffic police to read evidence of speeding vehicles and to punish speeding vehicles based on the evidence.

[0100] Therefore, the radar-video integrated automatic vehicle speeding detection system 100 provided in this application deploys radar-video integrated equipment at intersections, connects the radar-video integrated equipment to an algorithm server, connects the algorithm server to a data storage server, and connects the data storage server to an enforcement platform. When a vehicle speeds at an intersection, the enforcement platform displays speed, license plate, image, distance, time, and other evidence of speeding, providing reliable evidence for traffic management.

[0101] In the various embodiments described above, the specific order or hierarchy of steps in the disclosed process is an example of an exemplary method. Based on design preferences, it should be understood that the specific order or hierarchy of steps in the process may be rearranged without departing from the scope of this disclosure. The appended method claims provide elements of various steps in an exemplary order and are not intended to limit the scope to a specific order or hierarchy.

[0102] Those skilled in the art will also understand that the various illustrative logical blocks, modules, and steps listed in the embodiments of this application can be implemented by electronic hardware, computer software, or a combination of both. To clearly demonstrate the interchangeability of hardware and software, the functions of the various illustrative components, modules, and steps described above have been generally described. Whether such functionality is implemented through hardware or software depends on the specific application and the overall system design requirements. Those skilled in the art can implement the described functions using various methods for each specific application, but such implementation should not be construed as exceeding the scope of protection of the embodiments of this application.

[0103] The various illustrative logic blocks or modules described in the embodiments of this application can be implemented or operate the described functions using a general-purpose processor, digital signal processor, application-specific integrated circuit (ASIC), field-programmable gate array or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof. The general-purpose processor can be a microprocessor; alternatively, it can be any conventional processor, controller, microcontroller, or state machine. The processor can also be implemented using a combination of computing devices, such as a digital signal processor and a microprocessor, multiple microprocessors, one or more microprocessors combined with a digital signal processor core, or any other similar configuration.

[0104] The steps of the methods or algorithms described in the embodiments of this application can be directly embedded in hardware, a software module executed by a processor, or a combination of both. The software module can be stored in RAM, flash memory, ROM, EPROM, EEPROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium in the art. Exemplarily, the storage medium can be connected to the processor so that the processor can read information from and write information to the storage medium. Optionally, the storage medium can also be integrated into the processor. The processor and storage medium can be housed in an ASIC, which can be housed in a user terminal. Optionally, the processor and storage medium can also be housed in different components of the user terminal.

[0105] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of this application. It should be understood that the above description is only a specific embodiment of this application and is not intended to limit the scope of protection of this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for automatic vehicle speeding detection based on a radar-visual integrated machine, characterized in that, include: Acquire radar detection data and video image data; Based on the radar detection data, obtain the timestamp information and vehicle distance information of the target vehicle for each frame; Multiple vehicle speeds are obtained based on the timestamp information and distance information of the target vehicle corresponding to adjacent frames; The speeds of the multiple vehicles are sorted to obtain multiple preprocessed vehicle speeds, and the average of the multiple preprocessed vehicle speeds is applied to obtain the target speed of the target vehicle; wherein, the speeds of the multiple preprocessed vehicle speeds are greater than the speeds of all other vehicles among the multiple vehicle speeds. If the target speed is greater than the threshold speed, the target vehicle is a speeding vehicle. Based on the radar detection data, the coordinates of the center point of the vehicle position in the radar coordinate system are obtained. Perform coordinate system matrix transformation on the vehicle position center point coordinates in the radar coordinate system to obtain the vehicle position center point coordinates in the image coordinate system. Based on the video image data, an image of a speeding vehicle centered on the vehicle's position center point in the image coordinate system is obtained. Vehicle identification is performed on the speeding vehicle image to obtain the vehicle information of the speeding vehicle. The speeding vehicle image corresponds to the same timestamp as the vehicle's position center point coordinate in the radar coordinate system.

2. The automatic vehicle overspeed detection method based on a radar-visual integrated machine according to claim 1, characterized in that, The step of performing coordinate system matrix transformation on the coordinates of the vehicle's center point to obtain the coordinates of the vehicle's center point in the image coordinate system includes: The coordinates of the vehicle's center point are transformed using a camera coordinate system matrix to obtain the coordinates of the vehicle's center point in the camera coordinate system. The coordinates of the vehicle's center point in the camera coordinate system are transformed by an image coordinate system matrix to obtain the coordinates of the vehicle's center point in the image coordinate system.

3. The automatic vehicle overspeed detection method based on a radar-visual integrated machine according to claim 1, characterized in that, The step of acquiring an image of a speeding vehicle centered on the vehicle's center point in the image coordinate system based on the video image data, and performing vehicle recognition on the speeding vehicle image to obtain vehicle information of the speeding vehicle includes: Based on the video image data, obtain an image of the speeding vehicle with a length of 300 to 500 pixels and a width of 200 to 400 pixels, centered on the vehicle's position center point in the image coordinate system. Based on the target detection algorithm, the target region of the speeding vehicle image is extracted to obtain the license plate image; The license plate image is identified using a license plate recognition algorithm to obtain the license plate information of the speeding vehicle.

4. The automatic vehicle overspeed detection method based on a radar-visual integrated machine according to claim 1, characterized in that, The step of obtaining the timestamp information and vehicle distance information of the target vehicle corresponding to each frame based on the radar detection data includes: According to the Kalman filter tracking algorithm, the radar detection data is used to track the target vehicle, and the timestamp information and distance information of the target vehicle corresponding to each frame are obtained.

5. The automatic vehicle overspeed detection method based on a radar-visual integrated machine according to claim 1, characterized in that, The automatic vehicle overspeed detection method based on the radar-visual integrated machine also includes: Traffic management is performed on the speeding vehicle based on the timestamp information, distance information, target speed, image of the speeding vehicle, and vehicle information.

6. A vehicle overspeed automatic detection system based on a radar-visual integrated machine, characterized in that, include: The data acquisition module is used to acquire radar detection data and video image data; The target vehicle information processing module is used to obtain the timestamp information and vehicle distance information of the target vehicle corresponding to each frame based on the radar detection data. The vehicle speed processing module is used to obtain multiple vehicle speeds based on the timestamp information and vehicle distance information of the target vehicle corresponding to adjacent frames; The target speed processing module is used to sort the multiple vehicle speeds to obtain multiple pre-processed vehicle speeds, and to perform average processing on the multiple pre-processed vehicle speeds to obtain the target speed of the target vehicle; wherein, the multiple pre-processed vehicle speeds are greater than the speeds of all other vehicles among the multiple vehicle speeds. The overspeed vehicle coordinate position processing module is used to determine if the target vehicle is an overspeed vehicle if the target speed is greater than a threshold speed, and to obtain the coordinates of the center point of the overspeed vehicle's position based on the radar detection data. The coordinate transformation module is used to perform coordinate system matrix transformation on the coordinates of the vehicle's center point to obtain the coordinates of the vehicle's center point in the image coordinate system. The speeding vehicle information acquisition module is used to acquire a speeding vehicle image centered on the vehicle's position center point coordinates in the image coordinate system based on the video image data, perform vehicle recognition on the speeding vehicle image, and obtain the vehicle information of the speeding vehicle.

7. The automatic vehicle overspeed detection system based on a radar-visual integrated machine according to claim 6, characterized in that, The coordinate transformation module includes: The camera coordinate system transformation module is used to perform a camera coordinate system matrix transformation on the coordinates of the vehicle's center point to obtain the coordinates of the vehicle's center point in the camera coordinate system. The image coordinate system transformation module is used to perform an image coordinate system matrix transformation on the coordinates of the vehicle's center point in the camera coordinate system to obtain the coordinates of the vehicle's center point in the image coordinate system.

8. The automatic vehicle overspeed detection system based on a radar-visual integrated machine according to claim 6, characterized in that, The speeding vehicle information acquisition module includes: The speeding vehicle image processing module is used to acquire, based on the video image data, an image of the speeding vehicle with a length of 300 to 500 pixels and a width of 200 to 400 pixels, centered on the vehicle's position center point in the image coordinate system. The license plate image processing module is used to extract the target region from the speeding vehicle image according to the target detection algorithm to obtain the license plate image; The license plate information processing module is used to identify the license plate image according to the license plate recognition algorithm to obtain the license plate information of the speeding vehicle.

9. The automatic vehicle overspeed detection system based on a radar-visual integrated machine according to claim 6, characterized in that, The target vehicle information processing module includes: The timestamp vehicle distance acquisition module is used to track target vehicles based on the radar detection data according to the Kalman filter tracking algorithm, and obtain the timestamp information and vehicle distance information of the target vehicle corresponding to each frame.

10. The automatic vehicle overspeed detection system based on a radar-visual integrated machine according to claim 6, characterized in that, The automatic vehicle overspeed detection system based on the radar-visual integrated machine also includes: The radar-visual integrated machine is set up at an intersection to collect radar detection data and video image data, and transmit them to the data acquisition module. A data storage server is used to store the timestamp information, vehicle distance information, target speed, image of the speeding vehicle, and vehicle information of the speeding vehicle. The management platform is used to manage traffic for the speeding vehicle based on the timestamp information, distance information, target speed, image of the speeding vehicle, and vehicle information.

Images