A method of backtracking a vehicle

Abstract

This application discloses a vehicle retrospective tracking method, comprising: if a preset image acquisition device acquires image information of a target vehicle at a preset time, then, on the driving trajectory of the target vehicle, obtain the traffic state of a target road segment between the target image acquisition device and the preset image acquisition device at the preset time; using the speed coefficient corresponding to the traffic state at the preset time, determine the equivalent speed value of the target vehicle on the target road segment; calculate the time when the target vehicle passes the target image acquisition device based on the equivalent speed value and the preset time, to obtain the target time; query the target vehicle in the image information corresponding to the target time acquired by the target image acquisition device; after the target vehicle is found, retrospectively track the driving trajectory of the target vehicle based on the location information of the target image acquisition device.

Description

Technical Field

[0001] This application relates to the field of traffic management, and more particularly to a method for vehicle retrospective tracking. Background Technology

[0002] Highways are a crucial component of modern transportation systems. Compared to ordinary roads, highways have a higher accident severity and fatality rate. Therefore, tracking and tracing vehicles involved in traffic accidents or exhibiting abnormal driving behavior on highways is essential for preventing accidents, analyzing their causes, and improving highway traffic safety.

[0003] The main sensing equipment for highway information consists of high-definition cameras, video cameras, and inductive loop detectors. Existing products all rely on image information stored by high-definition cameras / video cameras for vehicle identification and tracking.

[0004] There are two main methods for tracing vehicles on highways:

[0005] One approach is to rely on high-definition cameras at highway checkpoints to capture images of passing vehicles, and then use license plate recognition to create a database of passing vehicles for target vehicle lookup.

[0006] Another method is to rely on manual querying of video information, utilizing the spatial relationships between image acquisition devices, and searching for vehicles one by one in the records of the preceding image acquisition devices of the current image acquisition device.

[0007] The aforementioned vehicle tracing method suffers from low efficiency and low accuracy. Summary of the Invention

[0008] To address any of the aforementioned technical problems, embodiments of this application provide a vehicle tracking and tracing method, comprising:

[0009] If the preset image acquisition device acquires image information of the target vehicle at a preset time, then the traffic status of the target road segment between the target image acquisition device and the preset image acquisition device at the preset time is obtained on the driving trajectory of the target vehicle.

[0010] Using the speed coefficient corresponding to the traffic state at the preset time, determine the equivalent speed value of the target vehicle on the target road segment;

[0011] Based on the equivalent speed value and the preset time, the time it takes for the target vehicle to pass through the target image acquisition device is calculated to obtain the target time;

[0012] The target vehicle is queried from the image information corresponding to the target time acquired by the target image acquisition device;

[0013] After the target vehicle is located, its driving trajectory is traced back based on the location information of the target image acquisition device.

[0014] One of the above technical solutions has the following advantages or beneficial effects:

[0015] By determining the target road segment between the target image acquisition device and the preset image acquisition device on the target vehicle's driving trajectory, the spatial information of the image acquisition device is utilized. Then, by acquiring the traffic state of the target road segment at the preset time, the equivalent speed value of the target vehicle corresponding to the target road segment is determined, and the target time when the target vehicle passes through the target image acquisition device is determined. This achieves the purpose of searching for vehicles within a specific location and time range by utilizing the time information of the image acquisition device, thereby improving the vehicle recognition efficiency and accuracy and efficiently completing vehicle backtracking.

[0016] Other features and advantages of the embodiments of this application will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the embodiments of this application. The objects and other advantages of the embodiments of this application may be realized and obtained by means of the structures particularly pointed out in the description, claims and drawings. Attached Figure Description

[0017] The accompanying drawings are used to provide a further understanding of the technical solutions of the embodiments of this application and constitute a part of the specification. They are used together with the embodiments of this application to explain the technical solutions of the embodiments of this application and do not constitute a limitation on the technical solutions of the embodiments of this application.

[0018] Figure 1 A flowchart of the vehicle identification method provided in the embodiments of this application;

[0019] Figure 2 A flowchart illustrating the process of determining the traffic status of a target road segment at a preset time, as provided in an embodiment of this application;

[0020] Figure 3 A flowchart of the vehicle recognition method provided for the application example of this application. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the embodiments of this application will be described in detail below with reference to the accompanying drawings. It should be noted that, unless otherwise specified, the embodiments and features in the embodiments of this application can be arbitrarily combined with each other.

[0022] In the process of developing this application, a technical analysis of the relevant technologies was conducted, and it was found that the relevant technologies have at least the following problems, including:

[0023] Existing target vehicle backtracking methods, such as license plate recognition, have certain requirements for weather (such as rain, snow, and other severe weather), photography environment (such as tunnels and other severe environments), high-definition camera performance, and high-definition camera coverage. Furthermore, they use the same search weight for vehicles that appear at different times and locations, failing to effectively utilize the spatiotemporal information of the highway network.

[0024] In addition, among the existing methods for tracking target vehicles, those based on manual search are labor-intensive and have low accuracy.

[0025] Based on the above analysis, the embodiments of this application provide the following solutions, including:

[0026] Figure 1 A flowchart illustrating the vehicle retrospective tracking method provided in this application embodiment. Figure 1 As shown, the method includes:

[0027] Step 101: If the preset image acquisition device acquires the image information of the target vehicle at a preset time, then on the driving trajectory of the target vehicle, obtain the traffic status of the target road segment between the target image acquisition device and the preset image acquisition device at the preset time.

[0028] Among them, the driving trajectory is the driving trajectory of the target vehicle on the highway;

[0029] The target acquisition device can be a camera installed on the road;

[0030] In one exemplary embodiment, the target image acquisition device can be an image acquisition device on a road segment that the target vehicle has already traveled through; or, it can be an image acquisition device on a road segment that the target vehicle is about to travel through.

[0031] In one exemplary embodiment, the traffic state value includes at least one of: free-flowing, smooth, congested, and blocked.

[0032] Step 102: Using the speed coefficient corresponding to the traffic state at the preset time, determine the equivalent speed value of the target vehicle on the target road segment;

[0033] In an exemplary embodiment, a correspondence between each traffic state and a speed coefficient can be established in advance so that the speed coefficient corresponding to the traffic state at the preset time can be determined based on the correspondence.

[0034] In one exemplary embodiment, the speed equivalent value is determined based on the speed coefficient and the driving speed of the target vehicle at the location of the preset image acquisition device.

[0035] For example, if the speed of the target vehicle at the location of the preset image acquisition device is A, and the speed coefficient corresponding to the traffic state at the preset time is B, then the equivalent speed value v' of the target vehicle on the target road segment is A*B.

[0036] Step 103: Calculate the time it takes for the target vehicle to pass the target image acquisition device based on the equivalent speed value and the preset time, and obtain the target time;

[0037] Let's take the distance of the target road segment as S, the preset time as t, and the equivalent speed as v' as an example for explanation:

[0038] If the target image acquisition device is located on a road segment where the target vehicle has already traveled, then the target time t is calculated using the following formula. o ,include:

[0039]

[0040] If the target image acquisition device is located on the road segment through which the target vehicle is about to travel, then the target time t is calculated using the following formula. o ,include:

[0041]

[0042] Step 104: Query the target vehicle from the image information corresponding to the target time acquired by the target image acquisition device.

[0043] Step 105: After the target vehicle is located, the driving trajectory of the target vehicle is traced back based on the location information of the target image acquisition device.

[0044] Specifically, the image acquisition devices in the upstream road segment where the target image acquisition device is located are determined to obtain the upstream devices. From these upstream devices, the device that acquired the image information of the target vehicle is selected to obtain the target upstream device. Then, based on the location information of the target image acquisition device and the target upstream device, the driving trajectory of the target vehicle is generated to obtain vehicle tracing information; or,

[0045] The image acquisition devices in the downstream road segment where the target image acquisition device is located are determined to obtain downstream devices. From the downstream devices, the device that acquired the image information of the target vehicle is selected to obtain the target downstream device. Based on the location information of the target image acquisition device and the location information of the target downstream device, the driving trajectory of the target vehicle is generated to obtain vehicle tracking information.

[0046] Taking vehicle backtracking as an example, the time information collected by the current target image acquisition device is taken as the current vehicle appearance time. The vehicle with matching image and time is searched in the image acquired by the image acquisition device at the previous position in space. This process is repeated to determine the time points and images of the target vehicle passing through all image acquisition devices, thereby realizing the trajectory backtracking of the target vehicle.

[0047] The method provided in this application embodiment determines the target road segment between the target image acquisition device and the preset image acquisition device on the driving trajectory of the target vehicle. It utilizes the spatial information of the image acquisition device, and then obtains the traffic state of the target road segment at the preset time to determine the equivalent speed value of the target vehicle on the target road segment. It also determines the target time when the target vehicle passes through the target image acquisition device, thereby utilizing the time information of the image acquisition device to achieve the purpose of searching for vehicles in a specific location and within a specific time range. This improves the vehicle recognition efficiency and accuracy, and efficiently completes vehicle backtracking.

[0048] The method provided in the embodiments of this application is described below:

[0049] In an exemplary embodiment, the traffic state of the target road segment at the preset time is obtained by means of:

[0050] Obtain sample data for the target road segment. The sample data includes at least two sets of traffic flow characterization data and traffic state values ​​corresponding to each set of traffic flow characterization data. Each set of traffic flow characterization data includes values ​​corresponding to at least two characterization parameters. The characterization parameters include at least one of flow rate, speed, and time occupancy rate.

[0051] Cluster analysis is performed on the sample data to calculate the baseline value of each characterization parameter corresponding to different traffic states, and the cluster centers corresponding to different traffic states are obtained.

[0052] The traffic flow characterization data of the target road segment at a preset time is used as the data to be analyzed, and the membership degree between the data to be analyzed and the cluster centers corresponding to different traffic states is determined.

[0053] Based on the membership degree, the value of the traffic state corresponding to the data to be analyzed is determined, and the traffic state of the target road segment at a preset time is obtained.

[0054] Figure 2 This is a flowchart illustrating the process of determining the traffic status of a target road segment at a preset time, as provided in an embodiment of this application. Figure 2As shown, traffic flow characteristics over a period of time are used as historical data. The Adaptive Fuzzy C-means (AFCM) algorithm is used to divide the data into four categories, and the membership degree of real-time traffic flow data to the cluster center is calculated to determine the current traffic status.

[0055] The AFCM algorithm is based on the Fuzzy C-means (FCM) algorithm, with the addition of weight W. k It is implemented, where the iteration method remains unchanged.

[0056] Among them, traffic condition measurement standard C is:

[0057]

[0058] In this matrix, the first column represents the average traffic flow per minute (Veh / 5min), the second column represents the vehicle speed on the road (km / h), and the third column represents the road's time occupancy rate (%). The first row represents the cluster centers corresponding to the road's smooth traffic condition, the second row represents the cluster centers corresponding to the stable traffic condition, the third row represents the cluster centers corresponding to the congested traffic condition, and the fourth row represents the cluster centers corresponding to the blocked traffic condition. The nine elements in the matrix represent the clustering results.

[0059] The traffic status is determined by calculating the membership degree between traffic flow parameter data and different types of cluster centers within C.

[0060] Using the above cluster analysis method to determine the traffic status at a preset time can improve the accuracy of the determination operation.

[0061] In an exemplary embodiment, a clustering analysis objective function is used to obtain a baseline value corresponding to traffic states for any characterization parameter at different values. The expression of the clustering analysis objective function is as follows:

[0062]

[0063]

[0064] in:

[0065] X represents the dataset representing the parameters, where X = [x1, x2, ..., x...]. n ], where n represents the total number of data points in the dataset representing the parameter;

[0066] U represents the fuzzy membership degree of the characterization parameter with respect to traffic states at different values, where u ik Represents the k-th data point x kThe fuzzy membership degree corresponding to the i-th value of the traffic state, where c represents the total number of values ​​for the traffic state;

[0067] V represents the baseline value of the characterizing parameter, where v i This represents the baseline value corresponding to the i-th value of the characterization parameter in the traffic state;

[0068] w k Represents the k-th data point x k The weights;

[0069] p represents the prior hyperparameter;

[0070] m represents the number of clusters, and m is a positive integer. In this scheme, m = 4.

[0071] In one exemplary embodiment, the k-th data point x k weight w k It is obtained through the following calculation expressions, including:

[0072]

[0073] In an exemplary embodiment, the reference value corresponding to the i-th value of the characterization parameter in traffic state is obtained by the following calculation expression, including:

[0074]

[0075] In one exemplary embodiment, the k-th data point x k The fuzzy membership degree u corresponding to the i-th value of traffic state ik It is obtained through the following calculation expressions, including:

[0076]

[0077] In an exemplary embodiment, determining the target vehicle from the image information corresponding to the target time captured by the target image acquisition device includes:

[0078] Sort the suspicious vehicles in the image information from high to low confidence, and select at least two of the top-ranked suspicious vehicles as candidate vehicles.

[0079] The probability of each candidate vehicle being identified as the target vehicle is calculated using the following formula, and the target vehicle is determined from the candidate vehicles based on the probability, including:

[0080] P car =αP ReID -βf(|Δt|);

[0081] Among them, P carLet PReID represent the probability, PReID represent the confidence level of the candidate vehicle, Δt represent the difference between the actual time of the candidate vehicle's appearance and the target time, f represent the window function, and α and β are both adjustment coefficients.

[0082] The following uses a camera installed on a highway as an example to illustrate the method provided in the embodiments of this application.

[0083] Figure 3 A flowchart illustrating the vehicle recognition method provided for an application example of this application. (e.g.) Figure 3 As shown, the method includes:

[0084] Step A: Video-based vehicle driving feature extraction, including:

[0085] Step A1: Video data acquisition;

[0086] Specifically, the cameras on the highway are screened, and the video streams captured by the selected cameras are obtained.

[0087] Step A2, vehicle driving feature extraction, including:

[0088] Specifically, deep learning-based object detection and tracking algorithms are used to perform real-time vehicle detection on the acquired video data. Examples include vehicle object detection based on YOLO (You Look Once) v3, vehicle object tracking based on DeepSort, and vehicle speed estimation based on lane line detection.

[0089] Based on the vehicle detection results, obtain the vehicle information at the camera location, including vehicle color, model, passing time, and estimated vehicle speed, as detailed in Table 1.

[0090] Table 1 Vehicle Driving Information Table

[0091] Camera logo date Vehicle Identification Arrival Time Model color speed C001 11-08-2021 1 11:13:38 car red 80.5 C001 11-08-2021 2 11:13:40 car white 86.91964 C001 11-08-2021 3 11:13:43 car black 82.5 C001 11-08-2021 4 11:13:45 car white 79.14634 C001 11-08-2021 5 11:13:47 car black 82.5 C001 11-08-2021 6 11:13:48 truck yellow 74.88462 C001 11-08-2021 7 11:13:49 bus yellow 76.05469 C001 11-08-2021 8 11:13:50 car white 83.92241 C001 11-08-2021 9 11:13:52 car white 81.80672

[0092] Step A3: Create a collection of vehicle images

[0093] Specifically, from the moment a vehicle enters the detection frame area until it leaves the detection frame area, an image with the center of the vehicle's return frame as the midpoint is retained every 10 frames. The image content is represented by "image + number" in the table below. A set of vehicle images to be retrieved is established. See Table 2 for details.

[0094] Table 2 List of Vehicles to be Searched

[0095] Camera logo date Vehicle Identification Arrival Time Vehicle images C001 11-08-2021 1 11:13:38 Image 1 C001 11-08-2021 2 11:13:40 Image 2 C001 11-08-2021 3 11:13:43 Image 3 C001 11-08-2021 4 11:13:45 Image 4 C001 11-08-2021 5 11:13:47 Image 5

[0096] Step A4: Calculate the road traffic flow characterization parameters at different locations using vehicle driving characteristics;

[0097] Specifically, with a sampling period of 5 minutes, the traffic flow, time, and time occupancy of each camera section are calculated to obtain a traffic flow characterization data table, as shown in Table 3.

[0098] Table 3 Traffic Flow Characterization Data

[0099]

[0100] Step B, near real-time traffic condition analysis and prediction, includes:

[0101] Step B1: Perform outlier detection and normalization on the traffic flow representation data;

[0102] Specifically, outliers can be removed first, and the remaining data after outlier removal can be normalized to complete the data preprocessing.

[0103] Step B2: Using the processed traffic flow characterization data, cluster the data to generate traffic state discrimination criteria;

[0104] Specifically, by using AFCM to cluster traffic flow characterization parameters, the traffic state metric C can be obtained as follows:

[0105]

[0106] In this matrix, the first column represents the average traffic flow per minute (Veh / 5min), the second column represents the vehicle speed on the road (km / h), and the third column represents the road's time occupancy rate (%). The first row represents the cluster centers corresponding to the road's smooth traffic condition, the second row represents the cluster centers corresponding to the road's stable traffic condition, the third row represents the cluster centers corresponding to the road's congested traffic condition, and the fourth row represents the cluster centers corresponding to the road's blocked traffic condition.

[0107] Step B3: Utilize the target vehicle's driving characteristics and road traffic conditions to obtain short-term communication prediction results for the target vehicle;

[0108] Specifically, with the current road traffic flow, time, and occupancy rate at 198 (Veh / 5min), 64.90 (km / h), and 8.80 (%), the traffic condition can be judged to be stable. Based on the traffic condition at this location, the target vehicle's speed and spatial position, and the road, the time interval of its passage past the preceding camera can be deduced, as detailed in Table 4.

[0109] Table 4 Forecast Timeline

[0110]

[0111] Step C, specific vehicle trajectory tracking and playback based on local features and traffic state analysis, includes:

[0112] Step C1: Vehicle re-identification based on Tranformer neural network, calculate the confidence image between the suspicious vehicle and the target vehicle;

[0113] Specifically, a vehicle re-identification model is completed by preprocessing vehicle driving characteristics and training the model. The vehicle re-identification model is then used to obtain confidence images between the suspicious vehicle and the target vehicle. These confidence images can be images of the same vehicle at the same angle, images of the same vehicle at different angles, images of different vehicles at the same angle, or images of similar vehicles at the same angle.

[0114] Step C2: Calculate the re-identification confidence of all suspicious vehicles and the target vehicle.

[0115] Step C3: Utilize the short-term communication prediction results of the target vehicle to estimate the vehicle's cargo, thereby obtaining the actual driving location of the suspicious vehicle and thus obtaining the location confidence level.

[0116] Step C4: Determine the probability that the suspicious vehicle is the target vehicle based on the re-identification confidence and the location confidence.

[0117] Specifically, based on the confidence level of identification and the confidence level of location, the vehicle with the highest confidence level is selected as the suspicious vehicle of the target vehicle, as detailed in Table 5.

[0118] Table 5. Probability Table of Suspicious Vehicles

[0119]

[0120] As shown in Table 5, vehicle 3 is the image record of the target vehicle being searched under camera 014.

[0121] Step C5: Using the time and speed of vehicle 3's appearance as the target position and speed, repeat step B3 to calculate the estimated arrival time of the vehicle at the previous camera; using the target vehicle image as the target image, determine the probability of the suspicious vehicle being the target vehicle, as detailed in Table 6:

[0122] Table 6. Probability Table of Suspicious Vehicles

[0123]

[0124] As shown in Table 6, Vehicle 1 is the image record of the target vehicle under camera 013. Similarly, images of the target vehicle under all cameras can be found, thus enabling vehicle route tracking across multiple cameras.

[0125] The method provided in this application embodiment, based on image data collected by cameras on highways, and based on vehicle image judgment, makes full use of the spatiotemporal information of the cameras and road traffic flow information, and uses the speed information of the target vehicle and the preceding road traffic state information to predict the time interval of the vehicle passing the preceding camera, and performs high-weight search on vehicles in specific locations and time ranges, thereby improving the performance of specific vehicle identification and backtracking.

[0126] This application provides a storage medium storing a computer program, wherein the computer program is configured to execute the method described in any of the preceding descriptions when it runs.

[0127] This application provides an electronic device including a memory and a processor. The memory stores a computer program, and the processor is configured to run the computer program to perform the method described in any of the preceding descriptions.

[0128] It will be understood by those skilled in the art that all or some of the steps, systems, or apparatuses disclosed above, and their functional modules / units, can be implemented as software, firmware, hardware, or suitable combinations thereof. In hardware implementations, the division between functional modules / units mentioned above does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed collaboratively by several physical components. Some or all components may be implemented as software executed by a processor, such as a digital signal processor or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit (ASIC). Such software may be distributed on a computer-readable medium, which may include computer storage media (or non-transitory media) and communication media (or transient media). As is known to those skilled in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and can be accessed by a computer. Furthermore, it is well known to those skilled in the art that communication media typically contain computer-readable instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

Claims

1. A method for retrospective tracking of vehicles, comprising: If the preset image acquisition device acquires image information of the target vehicle at a preset time, then the traffic status of the target road segment between the target image acquisition device and the preset image acquisition device at the preset time is obtained on the driving trajectory of the target vehicle; wherein, the target image acquisition device is the image acquisition device on the road segment that the target vehicle has already traveled through. Using the speed coefficient corresponding to the traffic state at the preset time, determine the equivalent speed value of the target vehicle on the target road segment; Based on the equivalent speed value and the preset time, the time it takes for the target vehicle to pass through the target image acquisition device is calculated to obtain the target time; The target vehicle is queried from the image information corresponding to the target time acquired by the target image acquisition device; After locating the target vehicle, the vehicle's trajectory is traced back based on the location information of the target image acquisition device, including: Acquire confidence images between each suspicious vehicle and the target vehicle, calculate the re-identification confidence corresponding to the confidence image of each suspicious vehicle; and acquire the position confidence between the actual location of each suspicious vehicle at the target time and the location of the target image acquisition device. Based on the re-identification confidence and location confidence of the same suspicious vehicle, determine the probability that each suspicious vehicle is identified as the target vehicle; Based on the probability corresponding to each suspicious vehicle, the vehicle that is identified as the target vehicle is determined from the suspicious vehicles, and the image of the target vehicle is obtained from the image information of the target image acquisition device.

2. The method according to claim 1, characterized in that, The traffic status of the target road segment at the preset time is obtained through the following methods: Obtain sample data of the target road segment. The sample data includes at least two sets of traffic flow characterization data and traffic state values ​​corresponding to each set of traffic flow characterization data. Each set of traffic flow characterization data includes values ​​corresponding to at least two characterization parameters. Cluster analysis is performed on the sample data to calculate the baseline value of each characterization parameter corresponding to different traffic states, and the cluster centers corresponding to different traffic states are obtained. The traffic flow characterization data of the target road segment at a preset time is used as the data to be analyzed, and the membership degree between the data to be analyzed and the cluster centers corresponding to different traffic states is determined. Based on the membership degree, the value of the traffic state corresponding to the data to be analyzed is determined, and the traffic state of the target road segment at a preset time is obtained.

3. The method according to claim 2, characterized in that, The characterization parameters include at least one of flow rate, speed, and time occupancy.

4. The method according to claim 2, characterized in that, The traffic status values ​​include at least one of: smooth, stable, congested, and blocked.

5. The method according to any one of claims 2 to 4, characterized in that, The objective function of cluster analysis is used to obtain the baseline value of any characterization parameter corresponding to different traffic states. The expression of the objective function of cluster analysis is as follows: in: The dataset represents the parameters, where , The total number of data points in the dataset representing the parameters; The fuzzy membership degree represents the relationship between the characterization parameter and the traffic state at different values, where Indicates the first Data points In traffic conditions The fuzzy membership degree corresponding to each value. It equals the total number of possible traffic state values; Represents the baseline value of the characterization parameter, where This indicates the characterization parameter in the traffic state. The baseline value corresponding to each value; Indicates the first Data points The weights; p represents the prior hyperparameter; m represents the number of clusters, and m is a positive integer.

6. The method according to claim 5, characterized in that, No. Data points weight It is obtained through the following calculation expressions, including: 。 7. The method according to claim 5, characterized in that, Characteristic parameters in traffic state The baseline value corresponding to each value is obtained through the following calculation expression, including: 。 8. The method according to claim 5, characterized in that: No. Data points In traffic conditions The fuzzy membership degree corresponding to each value It is obtained through the following calculation expressions, including: 。 9. The method according to claim 1, characterized in that: The equivalent speed value is determined based on the speed coefficient and the speed of the target vehicle at the location of the preset image acquisition device.

Images