Traffic situation analysis method and device

Abstract

The present application discloses a traffic situation analysis method and device, which relates to the field of traffic data processing, including: identifying image data in the video data of each section collected by the electric police, and determining the traffic parameters of each section; in the confluence section and the confluence section downstream, when there are sections whose traffic parameters within a preset time period meet the preset congestion conditions, counting the change trend of the number of vehicles whose corresponding vehicle speeds do not reach the preset speed in two sections with the same license plate number within the preset time range; if the change trend meets the preset trend, then determining that the two sections with the same license plate number are congested sections of the expressway. Based on this, the analysis results of the traffic situation of the expressway are closer to the actual traffic conditions, and the present application can accurately analyze the congestion situation of each section of the expressway by analyzing the expressway in multiple sections, and the traffic situation analysis results obtained are more accurate.

Description

Technical Field

[0001] The present application relates to the technical field of traffic data processing, and more specifically, to a traffic situation analysis method and device. Background Art

[0002] In the urban road system, expressways are long-distance fast passages within the city, used to improve the efficiency and speed of vehicle traffic within the city. However, with the increase in the number of vehicles, the traffic volume on expressways has also gradually increased, especially during special time periods such as holidays, morning and evening rush hours, the speed of vehicles on the main line of expressways is usually maintained below 20km / h, which is quite different from the expected speed at the beginning of the expressway design.

[0003] At present, the traffic situation of the expressway can be obtained through neural network algorithms, short-term predictions, etc., to provide reference for traffic police or signal light indication systems, so that they can take matching response measures to improve the vehicle traffic capacity on the main line of the expressway. For example: the ramp passing data can be continuously obtained through radar, cameras and other devices, the traffic flow data of the entire ramp can be calculated, and a traffic flow model can be established based on historical traffic data. The traffic situation of the ramp can be estimated through the traffic flow model, so as to analyze the traffic situation of the expressway under the influence of the traffic situation of the ramp. The macro traffic flow model can also be used to analyze the parameters that characterize the traffic characteristics of the main line of the expressway, and the traffic situation can be predicted in the short term based on the daily variation characteristics of the traffic flow obtained by analysis.

[0004] However, the above-mentioned prediction of traffic situation using the changing characteristics of ramps or traffic flows in the previous few days ignores the actual traffic conditions of the main line of the expressway at the current moment. The traffic situation of the expressway obtained by analysis has a lag and cannot accurately reflect the actual traffic conditions of the main line of the expressway at the current moment, making it impossible for traffic police or signal light indication systems to adopt matching response measures.

[0005] Based on this, how to accurately obtain the traffic situation of the expressway in real time has become an urgent problem to be solved. Summary of the invention

[0006] In view of the above problems, the present application provides a traffic situation analysis method and device to achieve the purpose of obtaining the traffic situation of the expressway in real time and accurately. The specific scheme is as follows:

[0007] In a first aspect, the present application provides a traffic situation analysis method, which is applied to a processor of a traffic situation analysis system, wherein the traffic situation analysis system at least includes: an electric alarm set at each section of the main line of the expressway, wherein the sections at least include: a merging area section, a merging area upstream section, and a merging area downstream section, wherein the traffic situation analysis method includes:

[0008] Acquire the video data of the section collected by each of the electric police, each of the video data includes multiple frames of image data;

[0009] Identify each frame of the image data in each of the video data, and determine the traffic parameters of each section, wherein the traffic parameters at least include: vehicle density, vehicle speed, and license plate number of each vehicle;

[0010] Determine whether there is a section in the merging area section and the section downstream of the merging area where the traffic parameters within a preset time period meet a preset congestion condition, wherein the preset congestion condition at least includes: whether the vehicle density and the vehicle speed of the section reach corresponding preset thresholds within a preset time period;

[0011] In the case where there is a section in the merging area section and the section downstream of the merging area where the traffic parameters within the preset time period meet the preset congestion condition, counting the change trend of the number of vehicles whose corresponding vehicle speeds do not reach the preset speed in the two sections with the same license plate number within the preset time range;

[0012] When the change trend meets a preset trend, the two sections with the same license plate numbers are determined to be congested sections.

[0013] In a possible implementation, the method further includes:

[0014] When there is no section in the merging area section and the downstream section of the merging area where the traffic parameters within the preset time period meet the preset congestion conditions, it is determined that the traffic situation of the main line of the expressway is in a smooth state.

[0015] In a possible implementation, identifying each frame of the image data in each of the video data and determining the traffic parameters of each section includes:

[0016] The processing process for each of the video data includes:

[0017] Identify the vehicles in each frame of the image data in the video data, and obtain the location information and license plate number of each vehicle;

[0018] Determine the driving track of each target vehicle based on the target vehicles with the same license plate number in each frame of the image data and the position information of each target vehicle;

[0019] Determine the vehicle density and vehicle speed corresponding to the section corresponding to the video data based on the number of the target vehicles and the driving trajectory;

[0020] The vehicle density, the vehicle speed and the set of the license plate numbers of each target vehicle corresponding to the section are determined as traffic parameters corresponding to the section.

[0021] In a possible implementation, determining the driving trajectory of each target vehicle based on the target vehicles with the same license plate number in each frame of the image data and the position information of each target vehicle includes:

[0022] Calling the last updated target tracking model, predicting the position information of the target vehicle in the image data of the next frame according to the position information of the target vehicle in the image data of the current frame, and obtaining the predicted position information of the target vehicle in the next frame, the target tracking model is constructed according to the Kalman filter, and updating the parameters of the state covariance matrix and the Kalman gain matrix in the target tracking model according to the predicted position information of the target vehicle in the current frame and the position information in the image data of the next frame, so that the target tracking model can predict the position information of the vehicle in the next frame according to the input vehicle position information;

[0023] Acquire the position information of each of the vehicles in the next frame of the image data;

[0024] The vehicle that satisfies a preset matching condition between the position information and the predicted position information is taken as the target vehicle;

[0025] Based on the position information of the target vehicle in the next frame and the current frame respectively, the driving trajectory of the target vehicle between the current frame and the next frame is determined.

[0026] In a possible implementation, determining the vehicle density and vehicle speed corresponding to the section corresponding to the video data based on the number of the target vehicles and the driving trajectory includes:

[0027] The driving trajectories of all the target vehicles corresponding to the video data, wherein the number of the position information contained does not meet a preset number, are filtered out to obtain a target driving trajectory;

[0028] The vehicle density and vehicle speed of each lane of the section corresponding to the video data are determined according to the target driving trajectory and the regular lines of the expressway main line, wherein the regular lines at least include lane lines and speed measurement lines.

[0029] In a possible implementation, identifying the vehicle in each frame of the image data in the video data to obtain the location information and license plate number of each vehicle includes:

[0030] Converting the image data into a preset data format to obtain target image data;

[0031] A pre-trained vehicle recognition model is called to perform vehicle recognition processing on the target image data to obtain the position information and license plate number of each vehicle in the target image data; the vehicle recognition model is trained based on the picture sample obtained by decoding the electric police video sample as a feature value, and the position information and license plate number of each vehicle in the picture sample is used as a target value.

[0032] In a possible implementation, the method further includes:

[0033] In the case where there is a section in the merging area section and the section downstream of the merging area where the traffic parameter in the preset time period satisfies the preset congestion condition, determining the number of sections in the merging area section, the section upstream of the merging area section and the section downstream of the merging area where the traffic parameter in the preset time period satisfies the preset congestion condition;

[0034] If the number of sections that meet the preset congestion condition is one, then determining that the traffic situation of the main line of the expressway is slightly congested;

[0035] If the number of sections that meet the preset congestion condition is two, then the traffic situation of the main line of the expressway is determined to be moderately congested;

[0036] If the number of sections that meet the preset congestion condition is three, it is determined that the traffic situation of the main line of the expressway is severely congested.

[0037] In a possible implementation, the step of obtaining the video data of the section collected by each of the electric police includes:

[0038] Obtain the rtsp stream addresses of the electric alarms set at the junction section of the expressway main line and the ramp intersection, as well as the upstream section and downstream section of the junction section upstream and downstream of the ramp intersection;

[0039] Acquire video data from the rtsp stream address corresponding to each of the electric police, and obtain video data corresponding to the confluence area section, the confluence area upstream section, and the confluence area downstream section respectively;

[0040] Video decoding is performed on each of the video data to obtain each frame of image data in each of the video data.

[0041] In a possible implementation, the method further includes:

[0042] When the traffic situation of the expressway main line is any one of slight congestion, moderate congestion and severe congestion, the traffic situation of the expressway main line and the congested road section are sent to a traffic light control system.

[0043] The second aspect of the present application provides a traffic situation analysis device, which is applied to a processor of a traffic situation analysis system, wherein the traffic situation analysis system at least includes: an electric alarm set at each section of the main line of the expressway, wherein the sections at least include: a merging area section, a merging area upstream section, and a merging area downstream section, and the traffic situation analysis device includes:

[0044] A data acquisition unit, used to acquire the video data of the section collected by each of the electric police, each of the video data includes multiple frames of image data;

[0045] A parameter determination unit, used for identifying each frame of the image data in each of the video data, and determining the traffic parameters of each section, wherein the traffic parameters at least include: vehicle density, vehicle speed, and license plate number of each vehicle;

[0046] a congestion state determination unit, configured to determine whether there is a section in the merging area section and the downstream section of the merging area where the traffic parameters within a preset time period meet a preset congestion condition, wherein the preset congestion condition at least includes: whether the vehicle density and the vehicle speed of the section reach corresponding preset thresholds within a preset time period;

[0047] a trend statistics unit, for counting the change trend of the number of vehicles whose corresponding vehicle speeds do not reach a preset speed in two sections with the same license plate number within a preset time range when the judgment result of the congestion state judgment unit is yes;

[0048] The congested section determination unit is used to determine that the two sections with the same license plate number are congested sections when the change trend meets a preset trend.

[0049] The traffic situation analysis method provided in the embodiment of the present application identifies the traffic parameters of the confluence section, the confluence section downstream section, and the confluence section upstream section of the current expressway based on the video data of each section collected by the electric police in real time, and identifies the congestion status of the expressway in sections according to the traffic parameters of each section. If there is congestion on the expressway, vehicles will stay in the congested section and will not be able to leave the expressway for a long time. Based on this, when the embodiment of the present application identifies congestion on the expressway, it searches for two sections with the same license plate numbers according to the traffic parameters of each section, and there is a possibility of congestion in the section. Further, based on the changing trend of the number of vehicles staying in the two sections, the changing trend of the traffic flow of the two sections with the possibility of congestion is monitored to determine the congested section that causes congestion on the main line of the expressway.

[0050] Compared with the existing technology of predicting traffic conditions based on historical data, the present application performs traffic situation analysis based on the traffic information of the expressway actually collected, and its analysis results are closer to the actual traffic conditions. In addition, the present application analyzes the congestion situation of each section of the expressway by analyzing the expressway in multiple sections, and the obtained traffic situation analysis results are more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0052] Figure 1 A schematic diagram of the layout of electric police on an expressway provided in an embodiment of the present application;

[0053] Figure 2 A flow chart of a method for implementing traffic situation analysis provided in an embodiment of the present application;

[0054] Figure 3 A schematic diagram of a vehicle driving trajectory provided in an embodiment of the present application;

[0055] Figure 4 A schematic diagram of a process for determining the congestion status of an expressway provided in an embodiment of the present application;

[0056] Figure 5 A schematic diagram of a process for identifying a congested section of an expressway provided in an embodiment of the present application;

[0057] Figure 6 A schematic diagram of the structure of a traffic situation analysis device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0058] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0059] Expressways are two-way roads with a central divider and all entrances and exits controlled by grade separations. They serve large-scale, long-distance, and fast traffic in cities. Expressways should have smooth lines and be separated from ordinary roads, so that cars can travel safely and smoothly at a higher speed. Expressways can also be called express highways, and often involve intercity expressways and urban expressways.

[0060] Taking into account the needs of high-speed and safe passage on expressways, traffic management departments, navigation systems, etc. all need to grasp the traffic situation of expressways in a timely manner, but the existing traffic situation analysis methods cannot accurately analyze the traffic situation of expressways. To solve the above problems, an embodiment of the present application provides a traffic situation analysis method, which is applied to a processor of a traffic situation analysis system, wherein the processor can be a lightweight embedded device with video data processing and network transmission functions, such as a Raspberry Pi station. The processor is used to analyze and process the video data of the expressway uploaded by the electric police to obtain the traffic situation of the expressway, wherein the electric police is an electronic police, which is a device that integrates modern advanced technologies, including video detection technology, computer technology, modern control technology, communication technology, computer network and database technology, etc.

[0061] In order to improve the accuracy of the traffic situation analysis of the expressway, the embodiment of the present application implements segmented analysis of the main line of the expressway through the layout of the electric police. Figure 1 , a schematic diagram of the layout of electric alarms for expressways provided in an embodiment of the present application, wherein the main line of the expressway is divided based on the ramp entrance to obtain the upstream section of the confluence area, the confluence area section, and the downstream section of the confluence area. An electric alarm is set in each section to collect video data of the road section.

[0062] Furthermore, the police at each section uploads the collected video data to the processor in real time, and the processor performs traffic situation analysis on the received video data. For details, please refer to Figure 2 , a flow chart of a method for implementing traffic situation analysis provided in an embodiment of the present application is provided to illustrate the method for analyzing traffic situation, and the specific process includes:

[0063] Step S110, obtaining the video data of the section collected by each electric police.

[0064] It can be understood that wireless communication can be achieved between the electric police and the processor. Therefore, the video data can be actively sent to the processor by the electric police, or the processor can access the electric police to obtain it. In one possible implementation, the process of the processor obtaining the video data of each section can include: obtaining the RTSP stream address of the electric police set for the merging area section between the main line of the expressway and the intersection, and the upstream section of the merging area and the downstream section of the merging area upstream and downstream of the ramp intersection; obtaining the video data from the RTSP stream address corresponding to each electric police, and obtaining the video data corresponding to the merging area section, the upstream section of the merging area and the downstream section of the merging area respectively; decoding each video data to obtain each frame of image data in each video data.

[0065] First, the processor can query the rtsp stream address of the currently monitored intersection of the expressway and the ramp, as well as the electric police upstream and downstream of the intersection. Obtain the corresponding video data from the rtsp stream address of each electric police, and obtain the video data of the confluence section corresponding to the electric police at the intersection of the expressway and the ramp, and the video data of the upstream and downstream sections of the confluence area corresponding to the upstream and downstream of the intersection.

[0066] It can be understood that a video is composed of multiple frames of continuous images. The embodiment of the present application can use a decoder to decode each obtained video data to obtain each frame of image data in the video data.

[0067] Step S120, identifying each frame of image data in each video data, and determining the traffic parameters of each section.

[0068] By tracking the vehicles in each frame of the video, the traffic volume, vehicle density, driving speed, etc. of the section corresponding to the video data can be obtained. For example, by tracking the same vehicle in two images corresponding to different times, the driving speed of the vehicle can be determined based on the position of the vehicle in the two images. Based on this, the driving speed of each vehicle traveling in the section corresponding to the video data can be determined, and the average driving speed can be used as the driving speed or traffic flow rate of the current section. The vehicle density can be obtained by identifying the number of vehicles in each frame of the image, and the average number of vehicles in each frame of the video data can be used as the vehicle density of the section.

[0069] In an embodiment of the present application, deep learning models such as 3D convolutional neural networks (3D-CNN) or recurrent neural networks (RNN), target tracking algorithms, image recognition, etc. can be used to perform overall recognition of video data or frame-by-frame recognition of image data in the video data to obtain traffic parameters that can characterize the traffic conditions of the section corresponding to the video data, wherein the traffic parameters may include: vehicle density, vehicle speed, license plate number of each vehicle, traffic flow, vehicle space occupancy, average spacing between vehicle heads, etc.

[0070] In one possible implementation, the process of acquiring traffic parameters of an embodiment of the present application may include the following steps, wherein, taking the processing process of video data of a section as an example, it may include: identifying vehicles in each frame of image data in the video data, and obtaining the position information and license plate number of each vehicle; determining the driving trajectory of each target vehicle based on the target vehicles with the same license plate number in each frame of image data and the position information of each target vehicle; determining the vehicle density and vehicle speed corresponding to the section corresponding to the video data based on the number of target vehicles and the driving trajectory; and determining the vehicle density, vehicle speed and license plate number of each target vehicle corresponding to the section as the traffic parameter corresponding to the section.

[0071] For each vehicle in each frame of the video data of the embodiment of the present application, the position information of the same vehicle is connected to obtain the driving trajectory of the vehicle in the section. It is understandable that expressways with different numbers of lanes can accommodate different numbers of vehicles, and the more lanes there are, the more vehicles can pass. For example: in the upstream section of the confluence area of ​​a two-way 8-lane expressway, 80 vehicles can be accommodated to travel smoothly, while in the upstream section of the confluence area of ​​a two-way 4-lane expressway, the passage of 80 vehicles will cause congestion on the expressway.

[0072] Therefore, the embodiment of the present application can determine the driving speed of the vehicle based on the driving trajectory of each vehicle during the video monitoring time, and can also determine the number of lanes in the section, so as to facilitate lane-level traffic parameter monitoring. For example, the vehicle whose driving trajectory is in the first lane is identified, and the traffic parameters such as the traffic density and driving speed of the first lane are determined, so as to determine the traffic situation of the first lane of the section, thereby improving the precision of traffic situation analysis.

[0073] Optionally, a pre-trained vehicle recognition model can be called to identify each frame of image data in the video data. The specific process may include: converting the image data into a preset data format to obtain target image data; calling a pre-trained vehicle recognition model to perform vehicle recognition processing on the target image data to obtain the location information and license plate number of each vehicle in the target image data; the vehicle recognition model is trained based on the picture sample obtained by decoding the electric police video sample as a feature value, and the location information and license plate number of each vehicle in the picture sample is used as the target value.

[0074] First, a large amount of image data containing vehicles is collected in advance. The embodiment of the present application selects to collect a large amount of image data of intersections, urban expressways, highways, etc. collected by the police, and marks the location information and license plate number of the vehicle in each image data, which is used as a training sample for the vehicle recognition model. The image data is used as the input data of the vehicle recognition model to be trained, and the location information and license plate number of the marked vehicle corresponding to each image data are used as the output data of the vehicle recognition model to be trained, and the trained vehicle recognition model is tested and verified, etc., to obtain a vehicle recognition model with a high vehicle detection rate and good generalization ability under the image data of the expressway.

[0075] Furthermore, since the image data that the trained vehicle recognition model can recognize has format restrictions, for example, the currently trained vehicle recognition model requires the input image size to be no larger than 640×640, while the image decoded from the video data of the electric police is usually no less than 2 million pixels, the image needs to be preprocessed before it is recognized by the vehicle recognition model. For example, operations such as proportional scaling, edge filling, pixel normalization, etc. that maintain the aspect ratio are performed to convert the image data into target image data that meets the format requirements. Each target image data is input into the vehicle recognition model to identify the location information and license plate number of each vehicle in the section.

[0076] It is understandable that if the vehicle recognition model has no format requirements for the input image data, the step of converting the image data format can be omitted, and each frame of image data can be directly input into the vehicle recognition model for processing.

[0077] Furthermore, based on the position information of the vehicle in each frame of image, the driving trajectory of the target vehicle with the same license plate number in each frame of image data is determined. Optionally, the trajectory of the same vehicle in each frame of image data can be tracked based on the license plate number of the vehicle.

[0078] In another possible implementation, the driving trajectory of each vehicle can also be determined based on the position information of the vehicle in each frame of image. The process may include: calling the target tracking model updated last time, predicting the position information of the target vehicle in the next frame of image data based on the position information of the target vehicle in the current frame of image data, and obtaining the predicted position information of the target vehicle in the next frame. The target tracking model is constructed based on the Kalman filter, and the state covariance matrix and the Kalman gain matrix in the target tracking model are updated according to the predicted position information of the target vehicle in the current frame and the position information in the next frame of image data, so that the target tracking model can predict the position information of the vehicle in the next frame according to the input vehicle position information; obtain the position information of each vehicle in the next frame of image data; take the vehicle that meets the preset matching conditions between the position information and the predicted position information as the target vehicle; and determine the driving trajectory of the target vehicle between the current frame and the next frame based on the position information of the target vehicle in the next frame and the current frame respectively.

[0079] The video data contains multiple frames of image data. The embodiment of the present application uses a pre-trained prediction model to predict the position of any vehicle in the image data of the current frame, and obtain the predicted position information. The processor detects the position information of each vehicle in the next frame of image data, and matches the position information of each vehicle with the predicted position information. If the position information of one of the vehicles is close to the predicted position information, the vehicle at the position information is identified as the same vehicle as the vehicle in the current frame corresponding to the predicted position information. Furthermore, based on the position information of the vehicle in the current frame and the next frame, the driving trajectory of the vehicle in these two frames is determined. According to the above method, the image data of all frames in the video data are identified to obtain the complete driving trajectory of each vehicle in the section.

[0080] Reference Figure 3 , a schematic diagram of a vehicle driving trajectory provided in an embodiment of the present application illustrates the process of determining the above-mentioned driving trajectory. The video data collected by the electric police contains 20 frames of images, namely frame a, frame b, frame c..., image recognition and vehicle trajectory tracking are performed frame by frame in chronological order. The current frame is frame a, and the image recognition finds that frame a contains the position information of multiple vehicles. One of the position information is selected, such as marking the vehicle at the position coordinate M1 as vehicle A, inputting the position coordinate M1 into the target tracking model, predicting the position information of vehicle A in frame b, and obtaining the predicted position coordinate N1.

[0081] Furthermore, image recognition is performed on frame b to obtain the position coordinates of each vehicle in frame b, such as position coordinates M 2-1 、M 2-2 、M 2-3 、M 2-4 、M2-5 , each position coordinate represents a vehicle. Each position coordinate is matched with the predicted position coordinate N2, and the vehicles corresponding to the two positions with the closest position coordinate distance are regarded as the same vehicle. Figure 3 , M in all position coordinates of frame b 2-2 It is closest to the predicted position coordinate N1, so M 2-2 The vehicle at the current position is identified as vehicle A, and the position coordinates M1 of the current frame a and the position coordinates M of the next frame b are connected. 2-2 , determine the driving trajectory of vehicle A during this period.

[0082] Based on this, the target vehicle matching between frame a and frame b is completed. Furthermore, the above process is repeated with frame b as the current frame according to the time sequence to complete the target vehicle matching between frame b and frame c, and the vehicle's driving trajectory from frame b to frame c is determined. This can be deduced by analogy to obtain the complete driving trajectory of each vehicle in the cross-section of the video data.

[0083] Optionally, the target tracking model used to predict the position information of the vehicle in the next frame in the embodiment of the present application can be constructed using a Kalman filter algorithm, and a set of Kalman filter parameters are pre-designed to run in the constructed target tracking model, and the position information of the next frame is predicted based on the position information of the current frame. After the position information detected in the next frame of image data is recognized through image recognition, the Hungarian algorithm is used to match the predicted position information with the position information of the next frame, and the minimum cost matrix Cost_Matrix is ​​constructed to complete the position matching of the same vehicle in the previous and next frames. Then, the predicted position information and the actual position information of the next frame identified can also be used to correct and update parameters such as the state covariance matrix and the Kalman gain matrix in the Kalman filter algorithm in the target tracking model to obtain an updated target tracking model. The updated target tracking model is applied to the position information prediction of the next frame. The embodiment of the present application performs parameter correction on the target tracking model based on the position information obtained in real time, so that it can learn the changing rules of the vehicle's trajectory under the current road conditions of the expressway, improve the accuracy of the output predicted position information, avoid errors in the predicted position information, and cause target vehicle matching errors. By correcting the target tracking model in real time, the accuracy of the predicted position is improved, thereby improving the accuracy of the target vehicle matching, and obtaining a more accurate driving trajectory, thereby improving the accuracy of the traffic parameters determined based on the driving trajectory.

[0084] Furthermore, based on the driving trajectory of each vehicle obtained from the video data of the tracking section, the traffic parameters of the section are determined. It can be understood that if the collected video data is 2 minutes long and contains 20 frames of image data, and a vehicle enters the section in the last 10 seconds of the video, the vehicle is detected in the last frame or two frames, and a driving trajectory covering the entire section cannot be generated. The reference value for calculating parameters such as vehicle speed and traffic flow in the section is not high.

[0085] Therefore, the process of determining the vehicle density and vehicle speed of a section based on the driving trajectory and the number of vehicles in the section according to the embodiment of the present application may include: filtering out the driving trajectories of all target vehicles corresponding to the video data whose number of position information does not meet a preset number, to obtain the target driving trajectory; determining the vehicle density and vehicle speed of each lane of the section corresponding to the video data based on the target driving trajectory and the rule lines of the main line of the expressway, wherein the rule lines include at least: lane lines and speed measurement lines.

[0086] When the vehicle is driving normally in the section, the position information of the identified vehicle corresponding to almost every frame of image data changes. The position information of the vehicle in each frame of image data can be regarded as a trajectory point. The trajectory points of multiple frames of images are connected to form the driving trajectory of the vehicle. The more trajectory points a vehicle has, the longer its driving trajectory in the section is, which is more conducive to the analysis of the traffic situation of the section. If the number of trajectory points is less, it will affect the accuracy of the traffic situation analysis of the section. For example, the electric police will take pictures of non-motorized lanes or parking spaces beside the expressway. If a vehicle is parked on the non-motorized lane, the position information of the non-motorized vehicle remains unchanged, that is, there is only one trajectory point. When performing image analysis and trajectory tracking on the section, the vehicles parked on the non-motorized lane in the image will also be analyzed. The parked vehicle has only one trajectory point and its speed is zero. If the driving speed of the section is calculated together with the driving trajectory and driving speed of the vehicles on the expressway, the driving speed of the section will be lowered, resulting in the subsequent misjudgment of the section as a congested state.

[0087] Based on this, in order to improve the accuracy of the traffic parameters of the section, the embodiment of the present application needs to screen the driving trajectory of the target vehicle. Optionally, vehicles with a certain number of track points in the driving trajectory are selected, for example, the driving trajectory with less than 10 track points is eliminated, and the remaining driving trajectories with more than 10 track points are analyzed. Optionally, the vehicle speed of the vehicle corresponding to the driving trajectory is determined according to the trajectory length of the vehicle driving trajectory and the interval between the first and last track points, and the average of the vehicle speeds of the selected driving trajectory is used as the vehicle speed of the section; based on the total number of driving trajectories, the total number of vehicles in the section, the number of lanes, the cross-sectional area, etc., the vehicle density of the section is determined.

[0088] Optionally, in order to improve the precision of traffic situation analysis, the embodiments of the present application may also add lane lines, speed measurement light rule lines in the image data, for example, add identification lane lines, and summarize each selected driving trajectory into each train lane, calculate the vehicle speed and traffic volume corresponding to each lane, so that the subsequent analysis of the traffic situation of each section of the expressway can be accurate to the lane.

[0089] Furthermore, the selected driving trajectory is matched with the calculated vehicle speed, vehicle density, license plate number, etc., to obtain the traffic parameters of the section.

[0090] Step S130, determining whether there is a section in the merging area section and the section downstream of the merging area whose traffic parameters meet the preset congestion condition within the preset time period. If the determination result is yes, executing step S140; if the determination result is no, executing step S160, determining that the traffic situation of the expressway is unblocked.

[0091] Considering that congestion on the expressway generally occurs in the confluence area and downstream of the confluence area, and congestion in any section of the confluence area and downstream of the confluence area will inevitably lead to congestion in the section upstream of the confluence area, therefore, the embodiment of the present application prioritizes confirming whether there is congestion on the current expressway by judging the congestion conditions of the two sections of the confluence area and downstream of the confluence area.

[0092] The preset congestion conditions include at least: whether the vehicle density and vehicle speed of the section reach the corresponding preset thresholds within the preset time period. For example, in the video data collected by the police for 5 consecutive 1-minute periods, the lane-level vehicle density of the section is not less than 5, and the vehicle speed is not less than 15km / h. The preset congestion conditions can be determined by the traffic data of previous expressway congestion. If the traffic parameters of any section in the confluence section and the downstream section of the confluence section meet the preset congestion conditions, it can be determined that the current traffic situation of the expressway is congested.

[0093] Optionally, during actual driving, although a road is in a congested state, the degree of congestion is different, and the response measures to be implemented are also different. Therefore, in order to improve the accuracy of identifying the congested state, the embodiment of the present application can also further identify the degree of congestion of the expressway in the process of identifying the congested state. The specific process may include: in the merging area section and the downstream section of the merging area, when there are sections whose traffic parameters within a preset time period meet the preset congestion conditions, determine the number of sections in the merging area section, the upstream section of the merging area, and the downstream section of the merging area whose traffic parameters within the preset time period meet the preset congestion conditions; if the number of sections that meet the preset congestion conditions is one, determine that the traffic situation of the main line of the expressway is slightly congested; if the number of sections that meet the preset congestion conditions is two, determine that the traffic situation of the main line of the expressway is moderately congested; if the number of sections that meet the preset congestion conditions is three, determine that the traffic situation of the main line of the expressway is severely congested.

[0094] The embodiment of the present application evaluates the congestion degree of the expressway by the number of sections that meet the preset congestion conditions among the multiple sections of the expressway. The more sections that meet the preset congestion conditions, the higher the congestion degree of the expressway. Conversely, the fewer sections that meet the preset congestion conditions, the lower the congestion degree of the expressway. Based on this, the relationship between each section and the preset congestion condition is judged to accurately determine the congestion degree of the expressway.

[0095] However, if the traffic parameters of the merging area section and the section downstream of the merging area do not meet the preset congestion conditions within the preset time period, or even the traffic parameters of the merging area section, the section downstream of the merging area and the section upstream of the merging area do not meet the preset congestion conditions within the preset time period, then the traffic situation of the expressway is determined to be unobstructed.

[0096] Step S140, counting the change trend of the number of vehicles whose corresponding vehicle speeds do not reach a preset speed in two sections with the same license plate number within a preset time range.

[0097] Step S150: When the change trend meets the preset trend, it is determined that two sections with the same license plate number are congested sections.

[0098] When it is determined that the traffic situation on the expressway is congested, the embodiment of the present application continues to analyze the traffic parameters of each section of the expressway to determine the congested sections of the expressway and improve the accuracy of the traffic situation analysis, so as to facilitate the feedback of the traffic situation analysis results to the management end and improve the accuracy of the adjustment measures for the congested status of the expressway.

[0099] If there is congestion on the expressway, the vehicle will travel slowly, and may even stay in the congested section, unable to leave the expressway for a long time. For example, at the vehicle speed expected on the expressway, it only takes 1 minute for a vehicle to travel from the upstream of the confluence area to the downstream of the confluence area. However, due to the congestion on the road section, the vehicle stays on the expressway for a long time, and it takes 10 minutes to travel from the upstream of the confluence area to the downstream of the confluence area. Based on this, if a vehicle fails to pass between two or even three sections for a long period of time, it may be that the vehicle stays for a long time due to the congestion on the expressway. Therefore, the embodiment of the present application uses the situation where the same vehicle appears in at least two sections within a preset time range as the initial condition for judging the congestion state of the expressway.

[0100] If the same vehicle is identified to appear in at least two sections within a preset time range, the road section where the two sections are located is preliminarily determined to be a congested section, but there may also be situations such as vehicle parking or vehicle failure, so further identification is required. It can be understood that the characteristics of road congestion include: a decrease in the average vehicle speed and an increase in vehicle density, and whether a road is congested depends on the vehicle with the lowest speed on the road and the number of vehicles with the lowest speed. Based on this, the embodiment of the present application determines the trend of the number of vehicles with lower speeds in the two sections through the traffic parameters of each section within a preset time range. If the trend is that the number of vehicles with low speeds increases, it proves that the section is about to be congested or is already in a congested state, and the section where the two sections are located can be determined as a congested section of the expressway.

[0101] In summary, the traffic situation analysis method provided in the embodiment of the present application identifies the traffic parameters of the confluence section, the confluence section downstream section, and the confluence section upstream section of the current expressway based on the video data of each section collected by the electric police in real time, and identifies the congestion status of the expressway in sections according to the traffic parameters of each section. If there is congestion on the expressway, the vehicle will stay in the congested section and will not be able to leave the expressway for a long time. Based on this, when the embodiment of the present application identifies congestion on the expressway, it searches for two sections with the same license plate numbers according to the traffic parameters of each section, and there is a possibility of congestion in the section. Further, based on the changing trend of the number of vehicles staying in the two sections, the changing trend of the traffic flow of the two sections with the possibility of congestion is monitored to determine the congested section that causes congestion on the main line of the expressway.

[0102] Compared with the existing technology of predicting traffic conditions based on historical data, the present application performs traffic situation analysis based on the traffic information of the expressway actually collected, and its analysis results are closer to the actual traffic conditions. In addition, the present application analyzes the congestion situation of each section of the expressway by analyzing the expressway in multiple sections, and the obtained traffic situation analysis results are more accurate.

[0103] Next, through Figure 4 and Figure 5 , the above-mentioned traffic situation analysis method is exemplified. In this example, the traffic situation analysis method is implemented using electric police cameras and Raspberry Pi stations installed at the junction of the expressway and the ramp, as well as the upstream and downstream of the junction. The Raspberry Pi station collects the video data of the section captured by the electric police camera, and performs vehicle recognition and target tracking on the image data in the video data to obtain the traffic parameters of each section under the electric police. The traffic parameters may include: vehicle speed, lane-level vehicle density, etc.

[0104] The traffic parameters of each section under the electric alarm are integrated and processed to obtain the traffic situation of the expressway. Figure 4 , a flow chart of judging the congestion status of an expressway provided in an embodiment of the present application, the Raspberry Pi station collects traffic parameters of the upstream section, the downstream section and the merging area, and judges in turn whether each section meets the preset congestion condition. In this example, it is judged in turn whether "the lane-level density of the downstream section of the merging area for 5 consecutive 1-minute periods is ≥5, and the vehicle speed is ≤15km / h", "the lane-level density of the merging area for 5 consecutive 1-minute periods is ≥5, and the vehicle speed is ≤15km / h", and "the lane-level density of the upstream section of the merging area for 5 consecutive 1-minute periods is ≥5, and the vehicle speed is ≤15km / h". Whether the traffic situation of the expressway is congested is determined based on the judgment results.

[0105] Among them, if both the downstream section and the section of the merging area do not meet the preset congestion conditions, the traffic situation of the expressway is unobstructed; if any one of the downstream section of the merging area, the section of the merging area and the upstream section of the merging area meets the preset congestion conditions, the traffic situation of the expressway is slightly congested; if any two of the downstream section of the merging area, the section of the merging area and the upstream section of the merging area meet the preset congestion conditions, the traffic situation of the expressway is moderately congested; and if the downstream section of the merging area, the section of the merging area and the upstream section of the merging area all meet the preset congestion conditions, the traffic situation of the expressway is severely congested, that is, heavy congestion.

[0106] If the expressway is determined to be congested, further identify the congested sections of the expressway, refer to Figure 5, a flow chart of identifying congested sections of expressways provided in an embodiment of the present application, collects vehicle license plate information of each section, and determines whether the same vehicle has license plate information in at least two sections. If not, it proves that the vehicle has not passed, or has just entered or exited the expressway, and the information is too little to be integrated to identify congested sections. If the same vehicle has license plate information in at least two sections, it can be preliminarily determined that there may be congested sections, and it is determined whether the sections where the same vehicle exists are three sections. If not, it is determined whether the sections where the same vehicle exists are adjacent sections. If they are adjacent sections, there are only two possibilities, namely "the same vehicle exists in the upstream section of the confluence area and the confluence area" and "the same vehicle exists in the confluence area section and the downstream section of the confluence area".

[0107] Reference Figure 5 , because the judgment process for the congested road section is almost the same in the above two cases, this example takes "the same vehicle exists in the upstream section of the merging area and the merging area" as an example. If the same vehicle exists in the upstream section of the merging area and the merging area, the difference between the time when the license plate information is obtained in the upstream section of the merging area and the merging area section is determined, that is, the difference between the time when the vehicle enters the upstream section of the merging area and the time when the vehicle enters the merging area section, and it is determined whether the difference is greater than a set value, where the set value can be the time it takes to pass through the two sections at the expected speed of the expressway. If the difference is greater than the time, it proves that the speed of the vehicle fails to reach the expected speed of the expressway, and there may be congestion.

[0108] Furthermore, the number of slow-moving vehicles in the upstream section and the merging area section is determined, and it is judged whether the number of slow vehicles is greater than or equal to 60% of the traffic volume in the past 5 minutes, that is, whether the number of slow-moving vehicles in the upstream section and the merging area section shows an increasing trend compared with the traffic volume in the past 5 minutes. If so, it is determined that there is congestion between the upstream section and the merging area section, and the section where the upstream section and the merging area section are located is determined as a congested section.

[0109] If the two sections with the same license plate information are not adjacent sections, that is, the same license plate information exists in the upstream section of the merging area and the downstream section of the merging area, then it is determined whether the difference between the time when the license plate information is obtained by the upstream section of the merging area and the downstream section of the merging area is greater than the set value, that is, whether the difference between the time when the vehicle enters the upstream section of the merging area and the time when the vehicle enters the downstream section of the merging area is greater than the set value, and the set value can be set to the length of time it takes to pass from the upstream section of the merging area to the downstream section of the merging area at the expected speed of the expressway. If the difference is greater than the set value, it proves that the vehicle's driving speed does not reach the expected speed of the expressway, so the section between the upstream section of the merging area and the downstream section of the merging area may be congested. Furthermore, the number of slow-moving vehicles in the upstream section and the downstream section of the merging area is determined. If the number is ≥ 60% of the traffic flow in the past 5 minutes, that is, the number of slow-moving vehicles in the upstream section and the downstream section of the merging area shows an increasing trend compared with the traffic flow in the past 5 minutes, the road section where the upstream section and the downstream section of the merging area are located is determined to be a congested section.

[0110] In this example, ramp signal light control is used to alleviate the traffic load on the main line of the expressway and increase the demand for the main line's traffic capacity. Therefore, when the traffic state or traffic situation of the main line of the expressway is any one of light congestion, moderate congestion and heavy congestion, the traffic state of the main line of the expressway and the congested section are sent to the signal light control system, so that it can respond in time and adjust the light control plan according to the real-time traffic situation of the expressway, thereby improving the efficiency and accuracy of expressway load regulation.

[0111] In addition to monitoring the traffic flow in the intersection area, the embodiment of the present application also adds monitoring of the upstream section area data and the downstream section area data of the intersection area, and without adding the original electric police equipment of the expressway, the video stream is obtained and decoded based on the existing electric police to obtain image frames, and the vehicles in the image frames are detected through the deep learning detection model. After obtaining the vehicle position information, the vehicle position information uses a tracking algorithm to generate the vehicle driving trajectory on the time series, and calculates the traffic parameters under each section through the vehicle trajectory and the pre-set lane lines, counting lines, speed measurement lines and other regular lines. At the same time, relying on the built-in capture function of the electric police, the license plate information of each vehicle in each section is obtained, and the same vehicle in different sections is matched through the license plate information; after obtaining the traffic parameter information and the license plate information of the same vehicle, the decision model is used to calculate the current traffic situation of the expressway, so as to make corresponding countermeasures according to the traffic situation and improve the road traffic capacity; this invention has a high adaptability to the scene, and also provides strong support and technical guarantee for subsequent decision analysis.

[0112] The traffic situation analysis device provided in an embodiment of the present application is described below. The traffic situation analysis device described below and the traffic situation analysis method described above can be referenced to each other.

[0113] First, combine Figure 6 , a traffic situation analysis device of a processor applied to a traffic situation analysis system is introduced, wherein the traffic situation analysis system at least includes: an electric alarm set at each section of the main line of the expressway, and the sections at least include: a merging area section, a merging area upstream section and a merging area downstream section, such as Figure 6 As shown, the traffic situation analysis device may include:

[0114] The data acquisition unit 100 is used to acquire the video data of the section collected by each of the electric police, each of the video data includes multiple frames of image data;

[0115] The parameter determination unit 200 is used to identify each frame of the image data in each of the video data, and determine the traffic parameters of each section, wherein the traffic parameters at least include: vehicle density, vehicle speed, and license plate number of each vehicle;

[0116] The congestion state determination unit 300 is used to determine whether there is a section in the merging area section and the downstream section of the merging area where the traffic parameters within a preset time period meet a preset congestion condition, wherein the preset congestion condition at least includes: whether the vehicle density and the vehicle speed of the section reach corresponding preset thresholds within a preset time period;

[0117] A trend statistics unit 400 is used for counting the change trend of the number of vehicles whose corresponding vehicle speeds do not reach a preset speed in two sections with the same license plate number within a preset time range when the judgment result of the congestion state judgment unit is yes;

[0118] The congested section determination unit 500 is used to determine that the two sections with the same license plate number are congested sections when the change trend meets a preset trend.

[0119] In summary, the traffic situation analysis device provided in the embodiment of the present application identifies the traffic parameters of the confluence section, the confluence section downstream section, and the confluence section upstream section of the current expressway based on the video data of each section collected by the electric police in real time, and identifies the congestion status of the expressway in sections according to the traffic parameters of each section. If there is congestion on the expressway, the vehicle will stay in the congested section and will not be able to leave the expressway for a long time. Based on this, when the embodiment of the present application identifies congestion on the expressway, it searches for two sections with the same license plate number according to the traffic parameters of each section, and there is a possibility of congestion in the section. Further, based on the changing trend of the number of vehicles staying in the two sections, the changing trend of the traffic flow of the two sections with the possibility of congestion is monitored to determine the congested section that causes congestion on the main line of the expressway.

[0120] Compared with the existing technology of predicting traffic conditions based on historical data, the present application performs traffic situation analysis based on the traffic information of the expressway actually collected, and its analysis results are closer to the actual traffic conditions. In addition, the present application analyzes the congestion situation of each section of the expressway by analyzing the expressway in multiple sections, and the obtained traffic situation analysis results are more accurate.

[0121] In a possible implementation, the traffic situation analysis device may further include:

[0122] The unblocked state determining unit is used to determine that the traffic situation of the main line of the expressway is unblocked when the judgment result of the congested state judging unit 300 is no.

[0123] In a possible implementation, the parameter determination unit 200 includes:

[0124] A vehicle identification subunit, used to identify the vehicle in each frame of the image data in the video data, and obtain the position information and license plate number of each vehicle;

[0125] A trajectory determination subunit, configured to determine the driving trajectory of each target vehicle based on the target vehicles with the same license plate number in each frame of the image data and the position information of each target vehicle;

[0126] A trajectory parameter determination subunit, used to determine the vehicle density and vehicle speed corresponding to the section corresponding to the video data based on the number of the target vehicles and the driving trajectory;

[0127] The traffic parameter determination subunit is used to determine the vehicle density, the vehicle speed and the set of license plate numbers of each target vehicle corresponding to the section as the traffic parameter corresponding to the section.

[0128] In a possible implementation, the trajectory determination subunit includes:

[0129] A tracking model calling subunit is used to call the last updated target tracking model, predict the position information of the target vehicle in the image data of the next frame according to the position information of the target vehicle in the image data of the current frame, and obtain the predicted position information of the target vehicle in the next frame. The target tracking model is constructed according to the Kalman filter, and the state covariance matrix and the Kalman gain matrix in the target tracking model are updated according to the predicted position information of the target vehicle in the current frame and the position information in the image data of the next frame, so that the target tracking model can predict the position information of the vehicle in the next frame according to the input vehicle position information;

[0130] A next frame information acquisition subunit, used for acquiring the position information of each vehicle in the next frame of image data;

[0131] An information matching subunit, configured to take the vehicle that satisfies a preset matching condition between the position information and the predicted position information as the target vehicle;

[0132] The trajectory determination subunit is used to determine the driving trajectory of the target vehicle between the current frame and the next frame based on the position information of the target vehicle in the next frame and the current frame respectively.

[0133] In a possible implementation, the trajectory parameter determination subunit includes:

[0134] A track screening subunit, used for screening out the driving tracks of all the target vehicles corresponding to the video data, the driving tracks of which the amount of the position information contained does not meet a preset amount, to obtain a target driving track;

[0135] The lane parameter determination subunit is used to determine the vehicle density and vehicle speed of each lane of the section corresponding to the video data based on the target driving trajectory and the regular lines of the expressway main line, and the regular lines at least include: lane lines and speed measurement lines.

[0136] In a possible implementation, the vehicle identification subunit includes:

[0137] A format conversion subunit, used for converting the image data into a preset data format to obtain target image data;

[0138] The model recognition subunit is used to call a pre-trained vehicle recognition model to perform vehicle recognition processing on the target image data to obtain the position information and license plate number of each vehicle in the target image data; the vehicle recognition model is based on the picture sample obtained by decoding the electric police video sample as a feature value, and the position information and license plate number of each vehicle in the picture sample are trained as target values.

[0139] In a possible implementation, the traffic situation analysis device may further include:

[0140] a congestion section quantity determination unit, configured to determine the quantity of the congestion sections in the confluence area section, the confluence area upstream section, and the confluence area downstream section, where the traffic parameters in the preset time period meet the preset congestion condition, if there are sections in the confluence area section and the confluence area downstream section where the traffic parameters in the preset time period meet the preset congestion condition;

[0141] A light congestion state determination unit, configured to determine that the traffic situation of the main line of the expressway is light congestion if the number of sections satisfying the preset congestion condition is one;

[0142] a medium congestion state determining unit, configured to determine that the traffic situation of the expressway main line is medium congestion if the number of sections satisfying the preset congestion condition is two;

[0143] A severe congestion state determination unit is used to determine that the traffic situation of the main line of the expressway is severe congestion if the number of sections that meet the preset congestion condition is three.

[0144] In a possible implementation, the data acquisition unit 100 includes:

[0145] The address acquisition subunit is used to obtain the rtsp stream addresses of the electric alarms respectively set at the confluence section of the expressway main line and the ramp intersection, and the upstream section and downstream section of the confluence section upstream and downstream of the ramp intersection;

[0146] A video acquisition subunit is used to acquire video data from the rtsp stream address corresponding to each of the electric police, and obtain video data corresponding to the confluence area section, the confluence area upstream section and the confluence area downstream section respectively;

[0147] The video decoding subunit is used to perform video decoding on each of the video data to obtain each frame of image data in each of the video data.

[0148] In a possible implementation, the traffic situation analysis device may further include:

[0149] The traffic situation feedback unit is used to send the traffic situation of the main line of the expressway and the congested road section to the traffic light control system when the traffic situation of the main line of the expressway is any one of light congestion, moderate congestion and heavy congestion.

[0150] It is understandable that before using the technical solutions disclosed in the embodiments of the present disclosure, the types, scope of use, usage scenarios, etc. of the personal information involved in the present disclosure should be informed to the user and the user's authorization should be obtained in an appropriate manner in accordance with relevant laws and regulations.

[0151] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0152] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0153] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A traffic situation analysis method, characterized in that: A processor applied to a traffic situation analysis system, the traffic situation analysis system at least further comprising: an electric alarm set at each section of the expressway main line, the sections at least comprising: a merging area section, a merging area upstream section and a merging area downstream section, the traffic situation analysis method comprising: Acquire the video data of the section collected by each of the electric police, each of the video data includes multiple frames of image data; Identify each frame of the image data in each of the video data, and determine the traffic parameters of each section, wherein the traffic parameters at least include: vehicle density, vehicle speed, and license plate number of each vehicle; Determine whether there is a section in the merging area section and the section downstream of the merging area where the traffic parameters within a preset time period meet a preset congestion condition, wherein the preset congestion condition at least includes: whether the vehicle density and the vehicle speed of the section reach corresponding preset thresholds within a preset time period; In the case where there is a section in the merging area section and the section downstream of the merging area where the traffic parameters within the preset time period meet the preset congestion condition, counting the change trend of the number of vehicles whose corresponding vehicle speeds do not reach the preset speed in the two sections with the same license plate number within the preset time range; In the case where the change trend meets a preset trend, determining that the two sections with the same license plate number are congested sections; The processing process for the image data in each video data includes: Call the last updated target tracking model, predict the position information of the target vehicle in the image data of the next frame according to the position information of the target vehicle in the image data of the current frame, and obtain the predicted position information of the target vehicle in the next frame, the target tracking model is constructed according to the Kalman filter, and the state covariance matrix and the Kalman gain matrix in the target tracking model are updated according to the predicted position information of the target vehicle in the current frame and the position information in the image data of the next frame, so that the target tracking model can predict the position information of the vehicle in the next frame according to the input vehicle position information, and the target vehicle is the vehicle with the same license plate number in the image data of each frame; Acquire the position information of each of the vehicles in the next frame of the image data; The vehicle that satisfies a preset matching condition between the position information and the predicted position information is taken as the target vehicle; Determine the driving trajectory of the target vehicle between the current frame and the next frame based on the position information of the target vehicle in the next frame and the current frame respectively; Determining the vehicle density and vehicle speed corresponding to the section corresponding to the video data based on the number of the target vehicles and the driving trajectory includes: The driving trajectories of all the target vehicles corresponding to the video data, wherein the number of the position information contained does not meet a preset number, are filtered out to obtain a target driving trajectory; The vehicle density and vehicle speed of each lane of the section corresponding to the video data are determined according to the target driving trajectory and the regular lines of the expressway main line, wherein the regular lines at least include lane lines and speed measurement lines.

2. The traffic situation analysis method according to claim 1, characterized in that: Also includes: When there is no section in the merging area section and the downstream section of the merging area where the traffic parameters within the preset time period meet the preset congestion conditions, it is determined that the traffic situation of the main line of the expressway is in a smooth state.

3. The traffic situation analysis method according to claim 1, characterized in that: The step of identifying each frame of the image data in each of the video data and determining the traffic parameters of each section includes: The processing process for each of the video data includes: Identify the vehicles in each frame of the image data in the video data, and obtain the location information and license plate number of each vehicle; Determine the driving track of each target vehicle based on the target vehicles with the same license plate number in each frame of the image data and the position information of each target vehicle; Determine the vehicle density and vehicle speed corresponding to the section corresponding to the video data based on the number of the target vehicles and the driving trajectory; The vehicle density, the vehicle speed and the set of the license plate numbers of each target vehicle corresponding to the section are determined as traffic parameters corresponding to the section.

4. The traffic situation analysis method according to claim 3, characterized in that: The identifying of the vehicle in each frame of the image data in the video data to obtain the position information and license plate number of each vehicle includes: Converting the image data into a preset data format to obtain target image data; A pre-trained vehicle recognition model is called to perform vehicle recognition processing on the target image data to obtain the position information and license plate number of each vehicle in the target image data; the vehicle recognition model is trained based on the picture sample obtained by decoding the electric police video sample as a feature value, and the position information and license plate number of each vehicle in the picture sample is used as a target value.

5. The traffic situation analysis method according to claim 1, characterized in that: Also includes: In the case where there is a section in the merging area section and the section downstream of the merging area where the traffic parameter in the preset time period satisfies the preset congestion condition, determining the number of sections in the merging area section, the section upstream of the merging area section and the section downstream of the merging area where the traffic parameter in the preset time period satisfies the preset congestion condition; If the number of sections that meet the preset congestion condition is one, then determining that the traffic situation of the main line of the expressway is slightly congested; If the number of sections that meet the preset congestion condition is two, then determining that the traffic situation of the expressway main line is moderately congested; If the number of sections that meet the preset congestion condition is three, it is determined that the traffic situation of the main line of the expressway is severely congested.

6. The traffic situation analysis method according to claim 1, characterized in that: The step of obtaining the video data of the section collected by each electric police comprises: Obtain the rtsp stream addresses of the electric alarms set at the junction section of the expressway main line and the ramp intersection, as well as the upstream section and downstream section of the junction section upstream and downstream of the ramp intersection; Acquire video data from the rtsp stream address corresponding to each of the electric police, and obtain video data corresponding to the confluence area section, the confluence area upstream section, and the confluence area downstream section respectively; Video decoding is performed on each of the video data to obtain each frame of image data in each of the video data.

7. The traffic situation analysis method according to any one of claims 1 to 6, characterized in that: Also includes: When the traffic situation of the expressway main line is any one of slight congestion, moderate congestion and severe congestion, the traffic situation of the expressway main line and the congested road section are sent to a traffic light control system.

8. A traffic situation analysis device, characterized in that: A processor applied to a traffic situation analysis system, the traffic situation analysis system at least further comprising: an electric alarm set at each section of the expressway main line, the sections at least comprising: a merging area section, a merging area upstream section and a merging area downstream section, the traffic situation analysis device comprising: A data acquisition unit, used to acquire the video data of the section collected by each of the electric police, each of the video data includes multiple frames of image data; A parameter determination unit, used for identifying each frame of the image data in each of the video data, and determining the traffic parameters of each section, wherein the traffic parameters at least include: vehicle density, vehicle speed, and license plate number of each vehicle; a congestion state determination unit, configured to determine whether there is a section in the merging area section and the downstream section of the merging area where the traffic parameters within a preset time period meet a preset congestion condition, wherein the preset congestion condition at least includes: whether the vehicle density and the vehicle speed of the section reach corresponding preset thresholds within a preset time period; a trend statistics unit, for counting the change trend of the number of vehicles whose corresponding vehicle speeds do not reach a preset speed in two sections with the same license plate number within a preset time range when the judgment result of the congestion state judgment unit is yes; a congested road section determining unit, configured to determine that the two sections having the same license plate number are congested road sections if the change trend satisfies a preset trend; Wherein, the functional unit for processing the image data in each of the video data includes: A tracking model calling subunit is used to call the last updated target tracking model, and predict the position information of the target vehicle in the image data of the next frame according to the position information of the target vehicle in the image data of the current frame, so as to obtain the predicted position information of the target vehicle in the next frame. The target tracking model is constructed according to the Kalman filter, and the state covariance matrix and the Kalman gain matrix in the target tracking model are updated according to the predicted position information of the target vehicle in the current frame and the position information in the image data of the next frame, so that the target tracking model can predict the position information of the vehicle in the next frame according to the input vehicle position information. The target vehicle is a vehicle with the same license plate number in the image data of each frame; A next frame information acquisition subunit, used for acquiring the position information of each vehicle in the next frame of image data; An information matching subunit, configured to take the vehicle that satisfies a preset matching condition between the position information and the predicted position information as the target vehicle; A trajectory determination subunit, used to determine the driving trajectory of the target vehicle between the current frame and the next frame based on the position information of the target vehicle in the next frame and the current frame respectively; The functional unit for determining the vehicle density and vehicle speed corresponding to the section corresponding to the video data based on the number of the target vehicles and the driving trajectory includes: A track screening subunit, used for screening out the driving tracks of all the target vehicles corresponding to the video data, the driving tracks of which the amount of the position information contained does not meet a preset amount, to obtain a target driving track; The lane parameter determination subunit is used to determine the vehicle density and vehicle speed of each lane of the section corresponding to the video data based on the target driving trajectory and the regular lines of the expressway main line, and the regular lines at least include: lane lines and speed measurement lines.

Images