Intersection vehicle overflow parking detection method, device and medium

Abstract

The embodiment of the invention discloses an intersection vehicle overflow parking detection method and device and a medium, belongs to the technical field of intelligent traffic, and solves the problem of low intersection passing efficiency. Comprising the following steps: performing framing processing on obtained intersection video stream information corresponding to an intersection to be detected to obtain an image set containing t frames of intersection road images; intercepting a corresponding four-direction cross passage area in each frame of intersection road image to obtain a required area image set after cutting; wherein each frame of image in the required area image set is respectively calibrated with an overflow stop area. And performing target detection and target tracking on each frame of image in the required area image set to obtain vehicle information corresponding to the overflow stop area in each frame of image. And comparing the vehicle information corresponding to the overflow stop area with a preset overflow stop vehicle threshold value, and determining a vehicle overflow stop condition corresponding to the to-be-detected intersection based on a comparison result, so as to perform signal lamp regulation and control on the to-be-detected road condition based on the vehicle overflow stop condition.

Description

Technical Field

[0001] The present application relates to the field of intelligent transportation technology, and in particular to a method, device and medium for detecting excessive parking of vehicles at intersections. Background Art

[0002] In today's world of rapid technological development, target tracking technology, as a key breakthrough in the field of intelligence, is profoundly changing the face of many industries, among which the rapid development of smart medical care, smart transportation and drone technology is particularly significant. Especially in the field of smart transportation, with the acceleration of urbanization and the surge in the number of motor vehicles, the traditional traffic management model faces severe challenges, such as road congestion during peak hours in the morning and evening, and uneven spatial and temporal distribution of commuting demand, which have become increasingly prominent, seriously affecting the travel experience of urban residents and the overall operating efficiency.

[0003] As the concept of smart city construction was proposed, various methods of "urban road congestion control" were proposed one by one. For example, the traffic volume of different roads was estimated through Internet traffic data, geomagnetic coil induction, vehicle-mounted GPS positioning and other methods, and then the timing of traffic lights was adjusted; or the real-time electric police video stream data was used to calculate the queue length in each direction of the current intersection and then dynamically adjust the timing of traffic lights.

[0004] Although traffic statistics can analyze the traffic demand at different times in the past, and queue length calculation can obtain the traffic demand at the current moment, the overflow parking caused by downstream congestion and conflicts between pedestrians, non-motor vehicles and motor vehicles still affects the traffic capacity of the intersection, thus reducing the traffic efficiency of the intersection. Summary of the invention

[0005] The embodiments of the present application provide a method, device and medium for detecting overflow parking of vehicles at an intersection, which are used to solve the following technical problems: overflow parking caused by downstream congestion and conflicts between pedestrians, non-motor vehicles and motor vehicles affects the traffic capacity of the intersection, thereby reducing the traffic efficiency of the intersection.

[0006] The present application embodiment adopts the following technical solutions:

[0007] An embodiment of the present application provides a method for detecting overflow stop of vehicles at intersections. The method comprises: performing frame processing on the obtained intersection video stream information corresponding to the intersection to be tested, and obtaining an image set containing t frames of intersection road images. The four-way cross-traffic areas corresponding to each frame of the intersection road image are intercepted to obtain a cropped required area image set; wherein each frame image in the required area image set is respectively marked with an overflow stop area. Target detection and target tracking are performed on each frame image in the required area image set to obtain vehicle information corresponding to the overflow stop area in each frame image. The vehicle information corresponding to the overflow stop area is compared with the preset overflow stop vehicle threshold, and the vehicle overflow stop situation corresponding to the intersection to be tested is determined based on the comparison result, so as to adjust the traffic lights for the road condition to be tested based on the vehicle overflow stop situation.

[0008] The embodiment of the present application detects and tracks the vehicles in each frame image in real time to ensure the real-time grasp of the vehicle driving status. By marking the overflow stop area on the image and comparing the vehicle information with the preset overflow stop vehicle threshold, a quantitative assessment of the vehicle overflow stop situation is achieved, avoiding errors caused by subjective judgment. Traffic lights are regulated based on the real-time vehicle overflow stop situation, which can dynamically optimize traffic flow, improve the efficiency of crossing vehicles, reduce congestion, and improve the utilization rate of road resources.

[0009] In one implementation of the present application, after obtaining the cropped required area image set, the method also includes: in each frame image, determining the area where the forward traffic flow and the lateral traffic flow intersect as the overflow stop main area; and, in each frame image, determining the area where the forward traffic flow leaves the overflow stop main area and enters the forward downstream lane as the forward overflow stop departure area; and, in each frame image, determining the area that the forward traffic flow must pass through when entering the overflow stop main area from the forward upstream lane as the forward overflow stop starting area; and, in each frame image, determining the area where the lateral traffic flow leaves the overflow stop main area and enters the lateral downstream lane as the lateral overflow stop departure area; and, in each frame image, determining the area that the lateral traffic flow must pass through when entering the overflow stop main area from the lateral upstream lane as the lateral overflow stop starting area; wherein each area corresponds to a preset vehicle stop judgment time threshold and a vehicle stop judgment quantity threshold.

[0010] In one implementation of the present application, target detection and target tracking are performed on each frame image in the required area image set, specifically including: performing target detection on the required area image set through the YOLOv8m model to obtain a motor vehicle target detection frame list; wherein the motor vehicle target detection frame list includes at least one of the image frame number, the detected motor vehicle number, the horizontal coordinate of the center point of the detection frame, the vertical coordinate of the center point of the detection frame, the width value of the detection frame, and the height value of the detection frame; performing target tracking on the motor vehicles in the required area image set through the Deepsort model to obtain a motor vehicle target tracking information list; wherein the motor vehicle target tracking information list includes at least one of the image frame number, the motor vehicle tracking number, the horizontal coordinate of the center point of the detection frame, the vertical coordinate of the center point of the detection frame, the width value of the detection frame, and the height value of the detection frame.

[0011] In one implementation of the present application, target detection and target tracking are performed on each frame image in the required area image set to obtain vehicle information corresponding to the overflow stop area in each frame image, specifically including: based on target detection and target tracking, determining the driving path information of the current vehicle, and determining the entry status of the current vehicle based on the driving path information; based on target detection and target tracking, determining the time information and vehicle quantity information of the current vehicle in the overflow stop area; based on the entry status of the current vehicle, the time information and vehicle quantity information of the overflow stop area, obtaining the vehicle information corresponding to the overflow stop area.

[0012] In one implementation of the present application, the entry status of the current vehicle is determined based on the driving path information, specifically including: when the current vehicle passes through the forward overflow stop exit area, determining that the current vehicle is a forward entry vehicle; when the current vehicle passes through the side overflow stop starting area, determining that the current vehicle is a side entry vehicle.

[0013] In one implementation of the present application, the time information and vehicle quantity information of the current vehicle in the overflow stop area are determined, specifically including: when the current vehicle first appears in area m corresponding to the overflow stop area in the i-th frame image, the time of first appearance in area m is determined; wherein area m is one of the overflow stop main area, the forward overflow stop departure area, the forward overflow stop starting area, the lateral overflow stop departure area and the lateral overflow stop starting area; when the current vehicle still appears in area m in the r-th frame image, the time corresponding to the r-th frame image is determined, and the number of overflow stop vehicles in area m corresponding to the r-1-th frame image is determined.

[0014] In one implementation of the present application, the vehicle information corresponding to the overflow stop area is compared with the preset overflow stop vehicle threshold, and the vehicle overflow stop situation corresponding to the intersection to be tested is determined based on the comparison result, specifically including: when the current vehicle still appears in area m in the rth frame image, time difference processing is performed based on the time corresponding to the rth frame image and the time when it first appears in area m; the obtained time difference is compared with the vehicle stop judgment time threshold corresponding to area m; and the number of overflow stopped vehicles in area m corresponding to the r-1th frame image is increased to obtain the number of overflow stopped vehicles in area m corresponding to the rth frame image; the number of overflow stopped vehicles in area m corresponding to the rth frame image is compared with the vehicle stop judgment number threshold corresponding to area m, so as to determine the vehicle overflow stop situation corresponding to area m of the intersection to be tested based on the comparison result.

[0015] In one implementation of the present application, signal light control is performed on the road condition to be tested based on the vehicle overflow stop situation, specifically including: when the overflow stop main area and / or the forward overflow stop departure area is in the overflow stop state, issuing a reminder to reduce the forward green light timing duration; when the overflow stop main area and / or the lateral overflow stop departure area is in the overflow stop state, issuing a reminder to reduce the lateral green light timing duration; when the overflow stop main area is in the overflow stop state and the forward overflow stop departure area and the lateral overflow stop departure area are in a normal state, issuing a pedestrian timing plan adjustment reminder or a traffic accident warning reminder.

[0016] An embodiment of the present application provides a vehicle overflow detection device at an intersection, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can: perform frame processing on the acquired intersection video stream information corresponding to the intersection to be tested, and obtain an image set containing t frames of intersection road images; intercept the four-way cross-traffic areas corresponding to each frame of the intersection road image, and obtain a cropped required area image set; wherein each frame image in the required area image set is respectively marked with an overflow stop area; perform target detection and target tracking on each frame image in the required area image set, and obtain vehicle information corresponding to the overflow stop area in each frame image; compare the vehicle information corresponding to the overflow stop area with a preset overflow stop vehicle threshold, and determine the vehicle overflow stop situation corresponding to the intersection to be tested based on the comparison result, so as to control the traffic lights for the road condition to be tested based on the vehicle overflow stop situation.

[0017] A non-volatile computer storage medium provided by an embodiment of the present application stores computer executable instructions, wherein the computer executable instructions are configured to: perform frame processing on the obtained intersection video stream information corresponding to the intersection to be tested, to obtain an image set including t frames of intersection road images; intercept the four-way cross-traffic areas corresponding to each frame of the intersection road image, to obtain a cropped required area image set; wherein each frame image in the required area image set is respectively marked with an overflow stop area; perform target detection and target tracking on each frame image in the required area image set, to obtain vehicle information corresponding to the overflow stop area in each frame image; compare the vehicle information corresponding to the overflow stop area with a preset overflow stop vehicle threshold, and determine the vehicle overflow stop situation corresponding to the intersection to be tested based on the comparison result, so as to control the traffic lights for the road condition to be tested based on the vehicle overflow stop situation.

[0018] At least one of the above technical solutions adopted in the embodiments of the present application can achieve the following beneficial effects: The embodiments of the present application ensure the real-time grasp of the vehicle's driving status by detecting and tracking the vehicle in each frame image in real time. By marking the overflow stop area on the image and comparing the vehicle information with the preset overflow stop vehicle threshold, a quantitative assessment of the vehicle overflow stop situation is achieved, avoiding errors caused by subjective judgment. Traffic lights are regulated based on the real-time vehicle overflow stop situation, which can dynamically optimize traffic flow, improve the efficiency of crossing vehicles, reduce congestion, and improve the utilization rate of road resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor. In the drawings:

[0020] Figure 1 A flow chart of a method for detecting excessive parking of vehicles at intersections provided in an embodiment of the present application;

[0021] Figure 2 An overflow stop division diagram of a four-way cross-traffic area at an intersection provided in an embodiment of the present application;

[0022] Figure 3 A diagram showing the actual effect of an intersection vehicle overflow detection provided in an embodiment of the present application;

[0023] Figure 4 A schematic diagram of the structure of an intersection vehicle overflow detection device provided in an embodiment of the present application.

[0024] Reference numerals:

[0025] 200: intersection vehicle over-parking detection device, 201: processor, 202: memory. DETAILED DESCRIPTION

[0026] The embodiments of the present application provide a method, device and medium for detecting vehicle overflow at an intersection.

[0027] In order to enable those skilled in the art to better understand the technical solutions in this application, 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 only part of the embodiments of this application, not all of them. Based on the embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this application.

[0028] The technical solution proposed in the embodiment of the present invention is described in detail below with reference to the accompanying drawings.

[0029] Figure 1 A flow chart of a method for detecting excessive parking of vehicles at an intersection provided in an embodiment of the present application is shown in FIG. Figure 1 As shown, the method for detecting excessive parking of vehicles at intersections includes the following steps:

[0030] S101, performing frame processing on the obtained intersection video stream information corresponding to the intersection to be tested, to obtain an image set including t frames of intersection road images.

[0031] In one embodiment of the present application, the video stream information of the electric police camera at the intersection is obtained, and the video is frame-processed to obtain an image set of t frames of the intersection road image, X={X 1 ,X 2 ,..X i ..,X t}, where X i is the i-th frame of the intersection electric police video image, i∈{1,…,t}.

[0032] Furthermore, the embodiment of the present application can obtain the video stream information of the electric police camera at the intersection through the ONVIF (Open Network Video Interface Forum) protocol to analyze it.

[0033] S102, intercepting the four-directional cross-traffic areas corresponding to the road image of each frame of the intersection, and obtaining a clipped required area image set.

[0034] In one embodiment of the present application, the i-th frame of the intersection electric police video image X i The area where traffic crosses in the middle and fourth directions is intercepted to obtain the i-th frame of the intersection electric police video image after trimming Get the required area image set X ′ ={X ′ 1 ,X ′ 2 ,..X i ′ ,..X ′ t}. In which, each frame image in the required area image set is marked with an overflow stop area.

[0035] Specifically, the slicing operation in the C++ language opencv library is used to crop the police video image set X to obtain the cropped image set Use cv::equalizeHist in the C++ language opencv library to crop Perform contrast enhancement to obtain the required area image set, that is, the four-way cross-traffic area image set X ′ ={X ′ 1 ,X ′ 2 ,..X i ′ ,..X ′ t}.

[0036] In one embodiment of the present application, in each frame image, the area where the forward traffic flow and the lateral traffic flow meet is determined as the overflow stop main area. Also, in each frame image, the area where the forward traffic flow leaves the overflow stop main area and enters the forward downstream lane is determined as the forward overflow stop departure area. Also, in each frame image, the area that the forward traffic flow must pass through when entering the overflow stop main area from the forward upstream lane is determined as the forward overflow stop starting area. Also, in each frame image, the area where the lateral traffic flow leaves the overflow stop main area and enters the lateral downstream lane is determined as the lateral overflow stop departure area. Also, in each frame image, the area that the lateral traffic flow must pass through when entering the overflow stop main area from the lateral upstream lane is determined as the lateral overflow stop starting area. Wherein, each area corresponds to a preset vehicle stop judgment time threshold and a vehicle stop judgment quantity threshold.

[0037] Specifically, in the intersection of four directions, the image set X ′ The overflow stop judgment area is calibrated, and the overflow stop time threshold and overflow stop vehicle number threshold are set for each area. Figure 2 A four-way intersection overflow stop division diagram is provided in an embodiment of the present application, such as Figure 2As shown, ① overflow stop main area: the area where the forward traffic flow and the lateral traffic flow intersect from the perspective of the electric alarm; ② forward overflow stop departure area: the area where the forward traffic flow leaves the overflow stop main area and enters the downstream lane; ③ forward overflow stop starting area: the area that the forward traffic flow must pass through when entering the overflow stop main area from the upstream lane; ④ lateral overflow stop departure area: the area where the lateral traffic flow leaves the overflow stop main area and enters the downstream lane; ⑤ lateral overflow stop starting area: the area that the lateral traffic flow must pass through when entering the overflow stop main area from the upstream lane.

[0038] Furthermore, given the vehicle stop judgment time threshold t (unit: seconds) and the vehicle stop judgment number threshold n (unit: vehicle) in each area: overflow stop main area Forward overflow exit area Forward overflow stop start area Side overflow stop area Lateral overflow stop start area It should be noted that in the environment used in the embodiment of the present application, t=8 and n=4 are taken. In actual applications, appropriate adjustments can be made according to the different conditions of each intersection, and the embodiment of the present application does not limit this.

[0039] S103 , performing target detection and target tracking on each frame image in the required area image set to obtain vehicle information corresponding to the overflow stop area in each frame image.

[0040] In one embodiment of the present application, the target detection is performed on the required area image set by the YOLOv8m model to obtain a list of motor vehicle target detection frames; wherein the list of motor vehicle target detection frames includes at least one of the image frame number, the number of the detected motor vehicle, the horizontal coordinate of the center point of the detection frame, the vertical coordinate of the center point of the detection frame, the width value of the detection frame, and the height value of the detection frame. The target tracking is performed on the motor vehicles in the required area image set by the Deepsort model to obtain a list of motor vehicle target tracking information; wherein the list of motor vehicle target tracking information includes at least one of the image frame number, the motor vehicle tracking number, the horizontal coordinate of the center point of the detection frame, the vertical coordinate of the center point of the detection frame, the width value of the detection frame, and the height value of the detection frame.

[0041] Specifically, the YOLOv8m model in the embodiment of the present application will analyze each frame of the image, identify the motor vehicle in the image, generate a bounding box for each detected motor vehicle, and output a list of motor vehicle target detection boxes at the same time. DeepSORT (Deep Learning for Real-Time Multi-Object Tracking and Segmentation) is a target tracking method that combines deep learning and traditional SORT (Simple Online and Realtime Tracking) algorithms, which can effectively track multiple targets in complex scenes. The embodiment of the present application uses the motor vehicle target detection box list output by the YOLOv8m model as the input of the DeepSORT model. DeepSORT initializes the tracker using the detected motor vehicle information, then matches and updates the tracking information of each motor vehicle between consecutive frames, and outputs a list of motor vehicle target tracking information.

[0042] Furthermore, the YOLOv8m model is used to train the image set X ′ Perform target detection and obtain the motor vehicle target detection box list detection_list i =[i,j,x,y,w,h], where i represents the i-th frame, j represents the j-th detected motor vehicle, x is the horizontal coordinate of the center point of the detection frame, y is the vertical coordinate of the center point of the detection frame, w is the width of the detection frame, and h is the height of the detection frame.

[0043] Furthermore, the Deepsort model is used to sort the image set X ′ The motor vehicle in the target is tracked, and the motor vehicle target tracking information list is obtained: tracker_list i =[i,k,x,y,w,h], where i represents the i-th frame, k represents the tracking number of the vehicle, x represents the horizontal coordinate of the center point of the detection frame, y represents the vertical coordinate of the center point of the detection frame, w represents the width of the detection frame, and h represents the height of the detection frame.

[0044] In one embodiment of the present application, based on target detection and target tracking, the driving path information of the current vehicle is determined, and the entry status of the current vehicle is determined based on the driving path information. Based on target detection and target tracking, the time information and vehicle quantity information of the current vehicle in the overflow stop area are determined. Based on the entry status of the current vehicle, the time information and vehicle quantity information of the overflow stop area, the vehicle information corresponding to the overflow stop area is obtained.

[0045] Specifically, by tracking the movement trajectory of the vehicle in multiple frames, the vehicle's driving path can be constructed. When the current vehicle passes through the forward overflow stop exit area, the current vehicle is determined to be a forward entry vehicle. When the current vehicle passes through the side overflow stop start area, the current vehicle is determined to be a side entry vehicle.

[0046] Specifically, the embodiment of the present application uses vector cross multiplication to determine whether the vehicle numbered k has passed through the forward overflow stop starting area. If it has passed through this area, k is recorded as a forward-entering vehicle; similarly, it is determined whether the vehicle numbered k has passed through the side overflow stop starting area. If it has passed through this area, k is recorded as a side-entering vehicle.

[0047] Specifically, the embodiment of the present application first defines the boundaries of the forward overflow starting area and the lateral overflow starting area. Among them, these boundaries can be represented by a series of line segments, and each line segment can be represented by a vector to represent its direction. For example, for the forward overflow starting area, its entrance boundary is defined as a line segment pointing from outside the area to inside the area, and the direction vector of the line segment is the boundary vector. When determining whether the vehicle passes through the area, the vehicle's driving direction is cross-multiplied with the vector of the area boundary. If the result of the cross-multiplication meets the preset passing condition, it is considered that the vehicle has passed through the area.

[0048] Furthermore, in the embodiment of the present application, if the cross product result of the driving direction of the vehicle numbered k and the entrance boundary vector of the front overflow starting area meets the "passing" condition during the driving process, then k is recorded as a front-entering vehicle. Similarly, if the cross product result of the vehicle numbered k and the boundary vector of the side overflow starting area meets the "passing" condition, then k is recorded as a side-entering vehicle.

[0049] In one embodiment of the present application, when the current vehicle first appears in the area m corresponding to the overflow stop area in the i-th frame image, the time of first appearance in the area m is determined; wherein the area m is one of the overflow stop main area, the forward overflow stop departure area, the forward overflow stop start area, the lateral overflow stop departure area, and the lateral overflow stop start area. When the current vehicle still appears in the area m in the r-th frame image, the time corresponding to the r-th frame image is determined, and the number of overflow stop vehicles in the area m corresponding to the r-1-th frame image is determined.

[0050] Furthermore, the embodiment of the present application also uses vector cross multiplication to determine whether the vehicle numbered k appears in the overflow stop determination area m for the first time in the i-th frame. If it appears for the first time, record Where m = 1, 2, 3, 4, 5, representing one of the overflow stop main area, forward overflow stop departure area, forward overflow stop start area, side overflow stop departure area and side overflow stop start area, respectively. i The current time (unit: seconds).

[0051] Furthermore, the vector cross multiplication method is used to determine whether the vehicle numbered k still appears in area m in the rth frame image. If it still appears, the time corresponding to the rth frame image is determined, and the number of overflow stopped vehicles in area m corresponding to the r-1th frame image is determined.

[0052] S104: Compare the vehicle information corresponding to the overflow stop area with the preset overflow stop vehicle threshold to determine the vehicle overflow stop situation corresponding to the intersection to be tested, so as to control the traffic lights for the road condition to be tested based on the vehicle overflow stop situation.

[0053] In one embodiment of the present application, when the current vehicle still appears in area m in the rth frame image, a time difference processing is performed based on the time corresponding to the rth frame image and the time when the vehicle first appears in area m; the obtained time difference is compared with the vehicle stop judgment time threshold corresponding to area m. Also, the number of overflow stopped vehicles in area m corresponding to the r-1th frame image is increased to obtain the number of overflow stopped vehicles in area m corresponding to the rth frame image. The number of overflow stopped vehicles in area m corresponding to the rth frame image is compared with the vehicle stop judgment number threshold corresponding to area m, so as to determine the vehicle overflow stop situation corresponding to area m of the intersection to be tested based on the comparison result.

[0054] Specifically, the current vehicle still appears in region m in the rth frame image, and Then the number of overflow stop vehicles in area m is in If at the rth frame, Region m is recorded as overflow stop state.

[0055] Among them, t r is the time corresponding to the rth frame image; is the time when the vehicle first appears in the overflow stop determination area m in the i-th frame; is the vehicle stop judgment time threshold corresponding to area m; is the number of overflow stopped vehicles corresponding to region m in the rth frame; is the number of overflow stopped vehicles in area m corresponding to the r-1th frame image; The vehicle stop determination quantity threshold.

[0056] In one embodiment of the present application, when the overflow stop main area and / or the forward overflow stop departure area is in an overflow stop state, a reminder to reduce the forward green light timing duration is issued. When the overflow stop main area and / or the lateral overflow stop departure area is in an overflow stop state, a reminder to reduce the lateral green light timing duration is issued. When the overflow stop main area is in an overflow stop state, and the forward overflow stop departure area and the lateral overflow stop departure area are in a normal state, a reminder to adjust the pedestrian timing plan or a traffic accident warning reminder is issued.

[0057] Specifically, if the overflow stop main area and the forward overflow stop departure area are in an overflow stop state, it means that the current forward traffic flow at the intersection causes an overflow stop at the intersection, and the forward green light timing needs to be reduced. If the overflow stop main area and the lateral overflow stop departure area are in an overflow stop state, it means that the current lateral traffic flow at the intersection causes an overflow stop at the intersection, and the lateral green light timing needs to be reduced. If the overflow stop main area is in an overflow stop state, and the forward overflow stop departure area and the lateral overflow stop departure area are in a normal state, it means that there are other situations at the intersection that prevent traffic from entering the downstream lane, and this may be caused by pedestrians and non-motor vehicles crossing the road, traffic accidents, etc. It is necessary to adjust the pedestrian timing plan or send traffic police to direct traffic.

[0058] Figure 3 A practical effect diagram of an intersection vehicle overflow detection provided in an embodiment of the present application, such as Figure 3 As shown in the figure, after obtaining the video from the electric police camera at the intersection, it is framed and intercepted to obtain the required area image. The detected motor vehicles in the figure have been marked with rectangular frames. By detecting and tracking the marked vehicles, the vehicle information corresponding to the overflow stop area can be determined, and then the overflow stop status of the vehicles at the current intersection can be determined.

[0059] Figure 4 This is a schematic diagram of the structure of a vehicle overflow detection device at an intersection provided in an embodiment of the present application. Figure 4As shown, the intersection vehicle overflow detection device 200 includes: at least one processor 201; and a memory 202 that is communicatively connected to the at least one processor 201; wherein the memory 202 stores instructions that can be executed by the at least one processor 201, and the instructions are executed by the at least one processor 201 so that the at least one processor 201 can: perform frame processing on the obtained intersection video stream information corresponding to the intersection to be tested, and obtain an image set containing t frames of intersection road images; intercept the four-way cross-traffic areas corresponding to each frame of the intersection road image, and obtain a cropped required area image set; wherein each frame image in the required area image set is respectively marked with an overflow stop area; perform target detection and target tracking on each frame image in the required area image set, so as to obtain vehicle information corresponding to the overflow stop area in each frame image; compare the vehicle information corresponding to the overflow stop area with the preset overflow stop vehicle threshold, and determine the vehicle overflow stop situation corresponding to the intersection to be tested based on the comparison result, so as to control the traffic lights for the road condition to be tested based on the vehicle overflow stop situation.

[0060] A non-volatile computer storage medium provided by an embodiment of the present application stores computer executable instructions, wherein the computer executable instructions are configured to: perform frame processing on the obtained intersection video stream information corresponding to the intersection to be tested, to obtain an image set including t frames of intersection road images; intercept the four-way cross-traffic areas corresponding to each frame of the intersection road image, to obtain a cropped required area image set; wherein each frame image in the required area image set is respectively marked with an overflow stop area; perform target detection and target tracking on each frame image in the required area image set, to obtain vehicle information corresponding to the overflow stop area in each frame image; compare the vehicle information corresponding to the overflow stop area with a preset overflow stop vehicle threshold, and determine the vehicle overflow stop situation corresponding to the intersection to be tested based on the comparison result, so as to control the traffic lights for the road condition to be tested based on the vehicle overflow stop situation.

[0061] Each embodiment in this application is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device, equipment, and non-volatile computer storage medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0062] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the embodiments of the present application may have various changes and variations. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present application.

Claims

1. A method for detecting excessive parking of vehicles at an intersection, characterized in that: The method comprises: Perform frame processing on the obtained intersection video stream information corresponding to the intersection to be tested to obtain an image set containing t frames of intersection road images; The four-way cross-traffic areas corresponding to each frame of the road image at the intersection are intercepted to obtain a clipped required area image set; wherein each frame image in the required area image set is marked with an overflow stop area; Performing target detection and target tracking on each frame image in the required area image set to obtain vehicle information corresponding to the overflow stop area in each frame image; The vehicle information corresponding to the overflow stop area is compared with the preset overflow stop vehicle threshold, and the vehicle overflow stop situation corresponding to the intersection to be tested is determined based on the comparison result, so as to control the traffic lights of the road condition to be tested based on the vehicle overflow stop situation.

2. A method for detecting excessive parking of vehicles at intersections according to claim 1, characterized in that: After obtaining the cropped required area image set, the method further includes: In each frame image, the area where the front traffic flow and the side traffic flow meet is determined as the overflow stop main area; and, in each frame image, determining an area where the forward vehicle flow leaves the overflow stop main area and enters the forward downstream lane as a forward overflow stop departure area; And, in each frame image, determining the area that the forward vehicle flow must pass through from the forward upstream lane into the overflow stop main area as the forward overflow stop starting area; and, in each frame image, determining an area where the lateral vehicle flow leaves the overflow stop main area and enters the lateral downstream lane as a lateral overflow stop departure area; And, in each frame image, determining the area that the lateral vehicle flow must pass through from the lateral upstream lane to enter the overflow stop main area as the lateral overflow stop starting area; Among them, each area corresponds to a preset vehicle stop judgment time threshold and a vehicle stop judgment quantity threshold.

3. A method for detecting excessive parking of vehicles at intersections according to claim 1, characterized in that: The performing target detection and target tracking on each frame image in the required area image set specifically includes: Performing target detection on the required area image set through the YOLOv8m model to obtain a list of motor vehicle target detection frames; wherein the list of motor vehicle target detection frames includes at least one of the image frame number, the number of the detected motor vehicle, the horizontal coordinate of the center point of the detection frame, the vertical coordinate of the center point of the detection frame, the width value of the detection frame, and the height value of the detection frame; The motor vehicles in the image set of the required area are tracked by the Deepsort model to obtain a motor vehicle target tracking information list; wherein the motor vehicle target tracking information list includes at least one of the image frame number, the motor vehicle tracking number, the horizontal coordinate of the center point of the detection frame, the vertical coordinate of the center point of the detection frame, the detection frame width value and the detection frame height value.

4. The method for detecting excessive parking of vehicles at intersections according to claim 1, characterized in that: The performing target detection and target tracking on each frame image in the required area image set to obtain vehicle information corresponding to the overflow stop area in each frame image specifically includes: Based on the target detection and the target tracking, determining the driving path information of the current vehicle, so as to determine the driving state of the current vehicle based on the driving path information; Based on the target detection and the target tracking, determining time information and vehicle quantity information of the current vehicle in the overflow stop area; Based on the current vehicle entry status, time information and vehicle quantity information of the overflow stop area, vehicle information corresponding to the overflow stop area is obtained.

5. A method for detecting excessive parking of vehicles at intersections according to claim 4, characterized in that: The determining the current vehicle entry state based on the driving path information specifically includes: In the case where the current vehicle passes through the forward overflow stop exit area, determining that the current vehicle is a forward entering vehicle; In a case where the current vehicle passes through the lateral overflow stop starting area, it is determined that the current vehicle is a sideways-entering vehicle.

6. A method for detecting excessive parking of vehicles at intersections according to claim 4, characterized in that: The determining of the time information and the number of vehicles of the current vehicle in the overflow stop area specifically includes: When the current vehicle first appears in the area m corresponding to the overflow stop area in the i-th frame image, determine the time of first appearance in the area m; wherein the area m is one of the overflow stop main area, the forward overflow stop departure area, the forward overflow stop start area, the lateral overflow stop departure area and the lateral overflow stop start area; When the current vehicle still appears in the area m in the rth frame image, the time corresponding to the rth frame image is determined, and the number of overflow stopped vehicles in the area m corresponding to the r-1th frame image is determined.

7. A method for detecting excessive parking of vehicles at intersections according to claim 6, characterized in that: The step of comparing the vehicle information corresponding to the overflow stop area with a preset overflow stop vehicle threshold, and determining the vehicle overflow stop situation corresponding to the intersection to be tested based on the comparison result, specifically includes: In the case where the current vehicle still appears in the area m in the rth frame image, performing time difference processing based on the time corresponding to the rth frame image and the time when the current vehicle first appears in the area m; The obtained time difference is compared with the vehicle stop judgment time threshold corresponding to the area m; and, increasing the number of overflow-stopped vehicles in the area m corresponding to the r-1th frame image to obtain the number of overflow-stopped vehicles in the area m corresponding to the rth frame image; The number of overflow stopped vehicles in the area m corresponding to the rth frame image is compared with the vehicle stop judgment number threshold corresponding to the area m, so as to determine the vehicle overflow stop situation corresponding to the area m of the intersection to be tested based on the comparison result.

8. The method for detecting excessive parking of vehicles at intersections according to claim 2, characterized in that: The step of performing signal light control on the road condition to be tested based on the vehicle overflow stop condition specifically includes: When the overflow stop main area and / or the forward overflow stop exit area is in an overflow stop state, a reminder to reduce the forward green light timing is issued; When the overflow stop main area and / or the lateral overflow stop departure area is in an overflow stop state, a reminder to reduce the lateral green light timing is issued; When the overflow stop main area is in the overflow stop state, and the forward overflow stop departure area and the lateral overflow stop departure area are in the normal state, a pedestrian timing plan adjustment reminder or a traffic accident warning reminder is issued.

9. A vehicle over-parking detection device at an intersection, characterized in that: The device comprises a memory for storing computer program instructions and a processor for executing the program instructions, wherein when the computer program instructions are executed by the processor, the device is triggered to execute the method according to any one of claims 1 to 8.

10. A non-volatile computer storage medium storing computer executable instructions, characterized in that: The computer executable instructions can execute the method according to any one of claims 1 to 8.

Images