Gate traffic state identification method and device, electronic equipment and storage medium

Abstract

The invention discloses a gate traffic state recognition method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining traffic influence factors of a target gate, and recognizing core influence factors in the traffic influence factors; determining a data acquisition area of the target gate based on the core influence factors, and obtaining traffic image information in the data acquisition area; extracting traffic state features of the traffic image information according to a preset feature extraction rule; and generating a traffic jam identification result of the target gate according to the traffic state characteristics of different data acquisition areas. According to the embodiment of the invention, the use efficiency of the road and the toll station can be improved and optimized, the gate congestion time is reduced, the traffic smoothness is improved, and the user experience is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of wireless communication, and in particular to a gate traffic state recognition method and device, an electronic device and a storage medium. BACKGROUND

[0002] With the rapid growth of highway traffic volume, the congestion of highway toll station, a "bottleneck" point, is becoming increasingly serious. However, as a node connecting the highway to the social road, the operation efficiency of the toll station directly affects the image and economic benefit of the highway. Scientific and reasonable operation state evaluation of the toll station can not only dynamically monitor the operation trend of the toll station, but also provide an important basis for the road management party to develop congestion alleviation measures. As an important connecting node of the highway and the social road, the operation efficiency of the toll station is not only related to the vehicle passing speed and travel time, but also directly related to the economic benefit and social benefit of the highway. Once the toll station is congested, not only will the vehicle queuing time be prolonged and the travel cost be increased, but also traffic accidents may be caused, which poses a threat to road traffic safety. However, the existing research on the highway toll station mostly focuses on the improvement of the passing capacity and the optimization of the lane layout, and the research on the operation efficiency and congestion state evaluation of the toll station is relatively less. This leads to a lack of scientific and effective means to evaluate the congestion state of the toll station in actual management, and it is difficult to provide strong support for the road management party to develop targeted congestion alleviation measures. SUMMARY

[0003] The present application provides a gate traffic state recognition method and device, an electronic device and a storage medium, which can facilitate traffic flow management, improve the use efficiency of the road and the toll station, reduce the gate congestion time, improve the smoothness of passing, and improve the user experience.

[0004] According to an aspect of the present application, a gate traffic state recognition method is provided, wherein the method comprises:

[0005] acquiring traffic influencing factors of a target gate, and identifying core influencing factors in the traffic influencing factors;

[0006] determining a data collection area of the target gate based on the core influencing factors, and acquiring traffic image information in the data collection area;

[0007] extracting traffic state features of each traffic image information according to a preset feature extraction rule;

[0008] generating a traffic congestion recognition result of the target gate according to the traffic state features of different data collection areas.

[0009] According to another aspect of the present application, a gate traffic state recognition device is provided, wherein the device comprises:

[0010] a core factor module configured to acquire traffic influence factors of a target gate and identify core influence factors within the traffic influence factors;

[0011] an image acquisition module configured to determine a data acquisition area of the target gate based on the core influence factors and acquire traffic image information within the data acquisition area;

[0012] a feature extraction module configured to extract traffic state features of each of the traffic image information according to a preset feature extraction rule;

[0013] a state identification module configured to generate a traffic congestion identification result of the target gate according to the traffic state features of different data acquisition areas.

[0014] According to another aspect of the present application, an electronic device is provided, which comprises:

[0015] at least one processor; and

[0016] a memory in communication connection with the at least one processor; wherein

[0017] the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to execute any of the gate traffic state identification methods of the present application.

[0018] According to another aspect of the present application, a computer readable storage medium is provided, which stores computer instructions for enabling a processor to implement the gate traffic state identification method of any of the embodiments of the present application when executed.

[0019] The technical solution of the embodiments of the present application acquires traffic influence factors of a target gate, determines core influence factors within the traffic influence factors, determines a data acquisition area of the target gate based on the core influence factors, acquires traffic image information within the data acquisition area, extracts traffic state features within the traffic image information, and determines a traffic congestion identification result of the target gate according to the traffic state features of different data acquisition areas. The embodiments of the present application quantitatively analyze traffic influence factors to determine core influence factors affecting gate traffic, which can improve the accuracy of traffic state identification, predict and manage traffic flow based on traffic state features of data acquisition areas, facilitate optimization of the use efficiency of roads and toll stations, reduce congestion and waiting time, improve traffic smoothness, and improve user experience.

[0020] It is to be understood that the details described in this section are not intended to identify key or critical elements of the embodiments of the application or to limit the scope of the application. Other features of the present application will become apparent from the following description. BRIEF DESCRIPTION OF DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative effort based on these drawings.

[0022] Figure 1 is a flow chart of a gate traffic state recognition method provided by an embodiment of the present application;

[0023] Figure 2 is a flow chart of another gate traffic state recognition method provided by an embodiment of the present application;

[0024] Figure 3 is a flow chart of another gate traffic state recognition method provided by an embodiment of the present application;

[0025] Figure 4 is a flow chart of another gate traffic state recognition method provided by an embodiment of the present application;

[0026] Figure 5 is an application example diagram of a gate traffic state recognition method provided by an embodiment of the present application;

[0027] Figure 6 is an example diagram of a gate traffic state recognition method provided by an embodiment of the present application;

[0028] Figure 7 is an example diagram of an influence factor weight provided by an embodiment of the present application;

[0029] Figure 8 is a congestion index diagram of a toll station in an idle period provided by an embodiment of the present application;

[0030] Figure 9 is a congestion index diagram of a toll station in an afternoon period of a working day provided by an embodiment of the present application;

[0031] Figure 10 is a congestion index diagram of a toll station in an afternoon period of a Saturday and Sunday provided by an embodiment of the present application;

[0032] Figure 11 is a congestion index diagram of a toll station in an afternoon period of a holiday provided by an embodiment of the present application;

[0033] Figure 12 Fig. 6 is a structural schematic diagram of a gate traffic state recognition device provided by an embodiment of the present application;

[0034] Figure 13 Fig. 7 is a structural schematic diagram of an electronic device for implementing a gate traffic state recognition method according to an embodiment of the present application. DETAILED DESCRIPTION

[0035] In order to enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work should fall within the scope of protection of the present application.

[0036] It should be noted that the terms "first", "second", and the like in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily indicate a specific order or a chronological sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but can include other steps or units that are not clearly listed or inherent to the process, method, product or device.

[0037] Embodiment One

[0038] Figure 1 Fig. 1 is a flowchart of a gate traffic state recognition method provided by an embodiment of the present application. The embodiment can be applied to a scenario of automatic recognition of gate traffic state. The method can be executed by a gate traffic state recognition device, which can be realized in the form of hardware and / or software, and can be configured in a terminal device or a server. As shown in Fig. 1, the method comprises the following steps. Figure 1

[0039] Step 110: acquiring traffic influencing factors of a target gate, and identifying core influencing factors in the traffic influencing factors.

[0040] ​The target gate can be a highway toll gate, and the traffic influencing factor can be an index factor influencing the congestion state of the target gate. The traffic influencing factor can include, but is not limited to, the number of vehicles at the toll station, the number of lanes at the toll station, the area of the toll station, the length of the lane at the toll station, the width of the lane at the toll station, the detection efficiency of the toll station, the traffic efficiency of the toll station, the ETC usage rate, the vehicle queue length, the weather factor, the holiday factor, and the equipment failure rate, etc. The core influencing factor can be an index that has a core effect on the target gate among the traffic influencing factors. The core influencing factor can be determined based on a weighted scoring method, a pair-wise comparison method, an entropy weight method, a principal component analysis method, etc.

[0041] In the embodiment of the present application, the traffic influencing factors influencing the traffic state of the target gate can be determined. The traffic influencing factors can be obtained by analyzing the historical data of the target gate. The importance of each traffic influencing factor can be determined based on the obtained traffic influencing factors. Some of the traffic influencing factors are selected as core application factors. The importance of the traffic influencing factors can be determined by a weighted scoring method, a pair-wise comparison method, an entropy weight method, a principal component analysis method, etc.

[0042] Step 120, determining the data collection area of the target gate based on the core influencing factor, and obtaining the traffic image information in the data collection area.

[0043] The data collection area can be a spatial area where the target gate is located. The data collection area can be divided according to the core influencing factor. Different data collection areas can correspond to different core influencing factors. The correspondence between the core influencing factor and the data collection area can be configured according to the actual environment of the target gate. The traffic image information can be image data collected by an image collection device in the data collection area. The traffic image information can include image data collected by a camera, image data collected by a drone, and infrared image data collected by an infrared camera, etc.

[0044] In the embodiment of the present application, the spatial area where the target gate is located can be obtained. The spatial area can be divided into multiple data collection areas according to the core influencing factor. The image information in each data collection area can be collected by using an image collection device. The traffic image information can be obtained, which can include road images, traffic marking images, and vehicle images, etc. in the data collection area.

[0045] Step 130, extracting the traffic state features of each traffic image information according to a preset feature extraction rule.

[0046] The preset feature extraction rule can be a preconfigured feature extraction rule for identifying traffic image information, and can be implemented based on a convolutional neural network or a target tracking algorithm.

[0047] Specifically, the preconfigured preset feature extraction rule can be acquired, and the preset feature extraction rule can be called to extract the traffic state features related to the road traffic congestion state, such as roads, traffic markings and vehicles in each frame of traffic image information.

[0048] Step 140, generating a traffic congestion recognition result of the target gate according to the traffic state features of different data collection areas.

[0049] In the embodiment of the present application, the traffic state features can be statistically analyzed for different data areas, and the traffic congestion recognition result of the target gate can be determined according to the statistical analysis results of each traffic state feature. The statistical analysis method can include but is not limited to the number of vehicles per unit time, the number of vehicles waiting in a single road, the number of stopped vehicles per unit area, etc. When the statistical analysis result meets certain requirements, it can be determined that the traffic congestion recognition result of the target gate is congested or not congested.

[0050] In the embodiment of the present application, the traffic influencing factors of the target gate are acquired, the core influencing factors in the traffic influencing factors are determined, the data collection area of the target gate is determined based on the core influencing factors, the traffic image information in the data collection area is collected, the traffic state features in the traffic image information are extracted, and the traffic congestion recognition result of the target gate is determined according to the traffic state features of different data collection areas. The embodiment of the present application quantitatively analyzes the traffic influencing factors to determine the core influencing factors affecting the gate traffic, which can improve the accuracy of traffic state recognition, predict and manage traffic flow based on the traffic state features of the data collection area, optimize the use efficiency of roads and toll stations, reduce congestion and waiting time, improve traffic smoothness, and improve user experience.

[0051] Embodiment two

[0052] Figure 2 is a flowchart of another gate traffic state recognition method provided by the embodiment two of the present application. The embodiment of the present application can explain the identification of the core influencing factors and the determination process of the data collection area, as shown in Figure 2 The method provided by the embodiment of the present application specifically includes the following steps:

[0053] Step 210, extracting gate traffic data of at least one data source according to the target gate.

[0054] In this embodiment of the invention, gate traffic data related to the target gate can be extracted from different data sources. The gate traffic data may include data forms such as text, images, and videos, and the data sources may include proprietary systems, third-party systems, etc.

[0055] Step 220: Classify the traffic data at the gate into traffic influencing factors according to preset factors.

[0056] Among them, the preset factor classification can be a pre-configured classification requirement or classification rule for classifying the traffic data at the gate. The preset factor classification can be provided by experts or artificial intelligence systems.

[0057] In this embodiment of the invention, the traffic data at the gate can be filtered and classified according to preset factors, and the classification results corresponding to each preset factor classification can be used as traffic influencing factors.

[0058] Step 230: Call the hierarchical analysis rules to determine the weight values ​​of each traffic influencing factor, and select the traffic influencing factors with a threshold number of weight values ​​as the core influencing factors.

[0059] The hierarchical analysis rules can be data processing rules built based on the Analytic Hierarchy Process (AHP), and can exist in the form of configuration files or applications. Weight values ​​can be parameters reflecting the magnitude of the impact of different traffic influencing factors on the congestion situation of the target gate; these weight values ​​can be obtained based on the hierarchical analysis rules.

[0060] In this embodiment of the invention, the hierarchical analysis rules can be invoked to process the obtained traffic influencing factors to obtain the weight value of each traffic influencing factor. The traffic influencing factors can be arranged from largest to smallest or from smallest to largest according to the weight value. Based on the sorting of the traffic influencing factors, the traffic influencing factors with larger weight values ​​can be selected as the core influencing factors.

[0061] Step 240: Obtain a three-dimensional spatial model of the space where the target gate is located, and determine the data collection area corresponding to the core influencing factors within the three-dimensional spatial model.

[0062] Specifically, a three-dimensional spatial model of the target gate can be obtained, which can be based on a geographic system or a simulation system.

[0063] In the embodiment of the present application, the spatial information of the target gate can be acquired, and a three-dimensional space model is constructed according to the spatial information. For example, the three-dimensional space model of the target gate can be searched in a geographic system according to the gate coordinates or the gate name of the target gate, or the point cloud data of the target gate is collected based on a depth camera, and the three-dimensional space model of the target gate is constructed based on the point cloud data. The three-dimensional space model can be divided into different data collection areas according to the core influencing factors.

[0064] In step 250, a preset image collection device is called to collect traffic image information in each data collection area.

[0065] The preset image collection device can be a device for collecting traffic image information, and can include but is not limited to one or more of a camera, a high-definition camera, a depth camera, and an infrared camera.

[0066] Specifically, for each data collection area, the preset image collection device can be called to collect images of objects in the above data collection area. Different data collection areas can use different or the same preset image collection device for collection. Traffic image information corresponding to different data collection areas can be acquired, which can include a sending and receiving island, a railing device, a lane marking, a road, a vehicle, etc.

[0067] In step 260, traffic state features of each traffic image information are extracted according to a preset feature extraction rule.

[0068] In step 270, a traffic congestion recognition result of the target gate is generated according to the traffic state features of different data collection areas.

[0069] In the embodiment of the present application, the gate traffic data in different data sources of the target gate is acquired, the gate traffic data is classified and divided into different traffic influencing factors according to a preset factor, the weight values of the traffic influencing factors are determined based on an analytic hierarchy rule, a threshold number of traffic influencing factors are selected as core influencing factors according to the weight values, a three-dimensional space model of the space where the target gate is located is acquired, the corresponding data collection areas in the three-dimensional space model are determined according to the core influencing factors, traffic image information is collected in the data collection areas according to a preset image collection device, traffic state features of each traffic image information are extracted according to a preset feature extraction rule, and the traffic state features are processed into a traffic congestion recognition result. The embodiment of the present application can realize quantitative analysis of traffic influencing factors, accurately acquire core influencing factors affecting gate traffic, improve the accuracy of traffic state recognition, predict and manage traffic flow based on traffic state features of data collection areas, facilitate optimization of the use efficiency of roads and toll stations, reduce congestion and waiting time, improve smoothness of traffic, and improve user experience.

[0070] Further, on the basis of the above-mentioned embodiment, the method further comprises: obtaining a consistency test result of the weight values of the traffic influence factors, and determining that the consistency test result is passed.

[0071] In the embodiment, after the weight values of the traffic influence factors are determined, consistency verification can be performed on the weight values, and after the weight values pass the consistency verification, that is, the consistency test result is passed, the core influence factor can be determined in the traffic influence factors according to the weight values.

[0072] For example, after the weight values of the traffic influence factors are calculated by the analytic hierarchy process, CI=(maximum eigenvalue-n) / (n-1) can be obtained, where n is the order of the matrix, that is, CI=0.065. The processing result generated by the analytic hierarchy process is input into the SPSS tool to obtain a corresponding random consistency RI value. The consistency index CR value is calculated based on CR=CI / RI. The test principle of the consistency test is that the smaller the CR value is, the better the test effect is. When the CR value is less than 0.1, it can be judged that the weight values meet the consistency requirement, and the consistency test result is passed.

[0073] Embodiment three

[0074] Figure 3 is a flowchart of another gate traffic state recognition method provided in the embodiment three of the present application. The embodiment of the present application describes the extraction process of the traffic state features, and the embodiment of the present application provides a method, which specifically comprises the following steps: Figure 3

[0075] Step 310: obtaining traffic influence factors of a target gate and recognizing core influence factors in the traffic influence factors.

[0076] Step 320: determining a data collection area of the target gate based on the core influence factors and obtaining traffic image information in the data collection area.

[0077] Step 330: recognizing a total number of vehicles in a region, a number of vehicles in an enabled lane and a queue length in the enabled lane in the traffic image information as traffic state features according to a target detection algorithm.

[0078] The target detection algorithm can be an algorithm for detecting and tracking a target in an image, and the target detection algorithm can include but is not limited to a SORT algorithm, a YOLO algorithm and the like. The total number of vehicles in a region can be the total number of vehicles in all lanes in the data collection area, the number of vehicles in the enabled lane can be the total number of vehicles in each enabled lane, and the queue length in the enabled lane can be the length of the queue of vehicles in each enabled lane. The queue length in the enabled lane includes the sum of the length of the vehicle and the length of the distance between vehicles.

[0079] ​In the embodiment of the present application, the vehicle coordinates and the entrance number of each vehicle in the traffic image information can be recognized based on the target detection algorithm, and the lane coordinate range of the lane can be recognized, the vehicle coordinates can be compared with the lane coordinate range, and the number of vehicles in each lane can be counted; the vehicle type of each vehicle can be recognized according to the target detection algorithm, the queue length in the enabled lane can be counted for each lane, the sum of the number of vehicles in all the enabled lanes can be taken as the total number of vehicles in the region, and the total number of vehicles in the region, the number of vehicles in the enabled lane and the queue length in the enabled lane counted can be taken as the traffic state features.

[0080] Step 340, the buffer area and the total number of vehicles in the buffer area in the traffic image information are recognized as the traffic state features according to the target detection algorithm.

[0081] The buffer area can be the area of the data collection region, and the total number of vehicles in the buffer area can be the total number of all vehicles in the data collection region.

[0082] In the embodiment of the present application, the vehicle coordinates and the entrance number of each vehicle in the traffic image information can be recognized based on the target detection algorithm, the area of each data collection region in the traffic image information can be taken as the buffer area, the coordinate range of the data collection region can be compared with the vehicle coordinates of each vehicle to determine the data collection region to which each vehicle belongs, the total number of vehicles in the data collection region can be determined for each data collection region as the total number of vehicles in the buffer area, and the buffer area and the total number of vehicles in the buffer area counted can be taken as the traffic state features.

[0083] Step 350, the total area of the ramp and the total number of vehicles staying in the ramp in the traffic image information are recognized as the traffic state features according to the target detection algorithm.

[0084] The total area of the region can be the area of different ramps in the traffic image information, and the total area of the region can be determined based on the coordinate range of each ramp. The total number of vehicles staying can be the total number of vehicles staying in different ramps.

[0085] In the embodiment of the present application, the vehicle coordinates and the entrance number of each vehicle in the traffic image information can be recognized based on the target detection algorithm, and the coordinate range of each ramp can be recognized, the total area of the ramp can be determined according to the coordinate range of each ramp, the coordinate range of each ramp can be compared with the vehicle coordinates of each vehicle to determine the vehicles included in each ramp, the total number of vehicles in each ramp can be counted as the total number of vehicles staying, and the total area of the ramp and the total number of vehicles staying counted can be taken as the traffic state features.

[0086] Step 360, the traffic congestion recognition result of the target gate is generated according to the traffic state features of different data collection regions.

[0087] The embodiment of the present application acquires the traffic influencing factors of the target gate, determines the core influencing factors in the traffic influencing factors, determines the data collection area of the target gate according to the core influencing factors, acquires the traffic image information according to the data collection area, identifies the total number of vehicles in the area, the number of vehicles in the active lane, the queue length in the active lane, the buffer area, the total number of vehicles in the buffer area, the total area of the ramp and the total number of parked vehicles in the traffic image information according to the target detection algorithm as the traffic state features, and generates the traffic congestion recognition result of the target gate according to the traffic state features of different data collection areas. The embodiment of the present application can identify the traffic state features in the image data by using the deep learning framework, can realize the rapid acquisition of the traffic dynamics, is convenient for improving the accuracy of the traffic state recognition, can predict and manage the traffic flow based on the traffic state features of the data collection area, is convenient for optimizing the use efficiency of the road and the toll station, reduces the congestion and waiting time, improves the smoothness of the traffic, and improves the user experience.

[0088] Embodiment four

[0089] Figure 4 is the flowchart of another gate traffic state recognition method provided by the fourth embodiment of the present application, the embodiment of the present application describes the generation process of the traffic congestion recognition result, and the embodiment of the present application provides the method, which specifically includes the following steps: Figure 4

[0090] Step 410, acquiring the traffic influencing factors of the target gate, and identifying the core influencing factors in the traffic influencing factors.

[0091] Step 420, determining the data collection area of the target gate based on the core influencing factors, and acquiring the traffic image information in the data collection area.

[0092] Step 430, extracting the traffic state features of each traffic image information according to the preset feature extraction rule.

[0093] Step 440, acquiring the traffic index evaluation model and the index weight of each historical traffic influencing factor, wherein the traffic index evaluation model and the index weight are generated based on the historical traffic data.

[0094] The historical traffic data can be a traffic data set of different gates in the past period of time, and the historical traffic data can include but is not limited to the traffic data of the target gate. The traffic index evaluation model can be composed of evaluation information of different traffic influencing factors, and the traffic index evaluation model at least includes the optimal value and the worst value of each type of traffic influencing factor.

[0095] ​In the embodiment of the present application, the historical traffic data of different gates can be collected, the historical traffic data can be respectively subjected to index statistical analysis according to historical traffic influencing factors, the optimal value and the worst value of each type of historical traffic influencing factor can be obtained to construct a traffic index evaluation model, and the index weight can be determined according to the factor value of each type of historical traffic influencing factor. For example, the index weight of each historical traffic influencing factor can be obtained by processing each type of traffic influencing factor through entropy method, pair comparison method, principal component analysis method, etc.

[0096] Step 450: determining the optimal index value and the worst index value in the traffic index evaluation model according to the traffic influencing factor of the traffic state feature, and normalizing the traffic state feature according to the optimal index and the worst index.

[0097] In the embodiment of the present application, for each type of traffic state feature, the optimal index value and the worst index value corresponding to each type of historical traffic influencing factor can be obtained in the traffic index evaluation model according to the traffic influencing factor to which the traffic state feature belongs, the historical traffic influencing factor can be matched to the traffic influencing factor to which the traffic state feature belongs, and the traffic state feature can be normalized according to the optimal index value and the worst index value. For example:

[0098] wherein, represents the optimal index value of the i-th traffic state feature of the j-th traffic influencing factor, represents the worst index value of the i-th traffic state feature of the j-th traffic influencing factor, and respectively represent the optimal index value and the worst index value of the j-th traffic influencing factor.

[0099] Step 460: determining the optimal Euclidean distance and the worst Euclidean distance between the traffic state feature and the optimal index and the worst index respectively.

[0100] Specifically, the Euclidean distance between the traffic state feature and the optimal index can be determined as the optimal Euclidean distance, and the Euclidean distance between the traffic state feature and the worst index can be determined as the worst Euclidean distance.

[0101] Step 470: determining the distance sum of the optimal Euclidean distance and the worst Euclidean distance, and taking the ratio of the worst Euclidean distance to the distance sum as the closeness value of the traffic state feature.

[0102] In the embodiment of the present application, the distance sum of the optimal Euclidean distance and the worst Euclidean distance corresponding to each traffic feature value can be calculated, and the ratio of the worst Euclidean distance to the distance sum is taken as the closeness value of the traffic state feature.

[0103] ​​In step 480, a weighted average closeness value is determined based on the closeness value of each traffic state feature and the index weight, and a congestion condition of the weighted average closeness value is looked up in a preset congestion degree table as a traffic congestion recognition result.

[0104] Specifically, the closeness value and the index weight of each traffic state feature can be obtained, the product of the closeness value and the index weight of each traffic state feature can be determined, the average of each product can be taken as the weighted average closeness value of each traffic state feature, the congestion state conforming to the weighted average closeness value can be looked up in the preset congestion degree table, and the congestion state can be taken as the traffic congestion recognition result. The preset congestion degree table can save the congestion states corresponding to different weighted average closeness values.

[0105] In the embodiment of the application, the traffic influence factors of the target gate are obtained, the core influence factors in the traffic influence factors are determined, the data collection area of the target gate is determined based on the core influence factors, the traffic image information in the data collection area is collected, the traffic state features in the traffic image information are extracted, the traffic index evaluation model generated based on historical traffic data and the index weight of each traffic influence factor are obtained, the optimal index value and the worst index value corresponding to the traffic state features in the traffic index evaluation model are determined, the traffic state features are normalized according to the optimal index and the worst index, the optimal Euclidean distance and the worst Euclidean distance between the traffic state features are determined, the closeness value of the traffic state features is determined according to the optimal Euclidean distance and the worst Euclidean distance, the weighted average closeness value is determined based on the closeness value of each traffic state feature and the index weight, and the congestion condition of the weighted average closeness value is looked up in a preset congestion degree table as a traffic congestion recognition result. The embodiment of the application can realize objective quantification of evaluation criteria, eliminate subjective bias and objective bias in the traffic congestion evaluation process, and make the evaluation of gate traffic more scientific and reliable.

[0106] In some embodiments of the application, the traffic index evaluation model and the index weight of each traffic influence factor are obtained, including:

[0107] The historical traffic data are obtained, and the historical traffic data are classified and divided into different historical traffic influence factors according to preset factors. For each historical traffic influence factor, the historical traffic influence factor is standardized based on the maximum value and the minimum value in the historical traffic influence factor. The entropy value of each historical traffic influence factor is determined, and the difference coefficient corresponding to the entropy value is obtained. The ratio of each difference coefficient to all difference coefficients is taken as the index weight of the historical traffic influence factor. The maximum value and the minimum value of each historical traffic influence factor are saved as the traffic index evaluation model.

[0108] In the embodiment of the present application, the obtained historical traffic data can be classified and divided into different historical traffic influence factors according to preset factors, the maximum value and the minimum value in each type of historical traffic influence factor can be extracted, the factor values in each type of historical traffic influence factor can be standardized according to the maximum value and the minimum value, the entropy value of each type of historical traffic influence factor can be determined according to the entropy value method for each type of historical traffic influence factor after standardization, the difference coefficient of each type of historical traffic influence factor can be calculated according to the entropy value, for example, wherein, represents the difference coefficient of the i-th type of historical traffic influence factor, represents the entropy value of the i-th type of historical traffic influence factor. The ratio of each difference coefficient to all difference coefficients can be taken as the index weight of the historical traffic influence factor, and the maximum value and the minimum value of each historical traffic influence factor can be saved as a traffic index evaluation model.

[0109] Embodiment five

[0110] The embodiment of the present application provides a gate traffic state recognition method, referring to Figure 5 The method uses advanced intelligent monitoring and radar equipment to continuously and comprehensively track and identify the front field of the gate and the ramp before the station. Through deep learning and pattern recognition technology, the multi-dimensional information of passing vehicles can be filtered and recorded in real time combined with the real-time database of the expressway. Specifically, the system extracts data from 12 key influence factors such as the number of vehicles, the number of lanes, the lane width, and the traffic efficiency, and uses a complex multi-level analysis method to process the 12 key influence factors. Each data is accurately weighted and evaluated from multiple dimensions to assess the congestion status of the toll station. Then, according to the comprehensive evaluation score, the system will give a visual graph for early warning, and combined with the results of the advantage and disadvantage evaluation, it will provide a suitable preset relief scheme to optimize the traffic flow of the toll station. This method not only can realize accurate evaluation of congestion state, but also can adjust the traffic management strategy in real time through the intelligent decision support system, timely feedback the congestion state, effectively relieve the traffic pressure during peak hours, ensure the efficient and safe passing of vehicles through the toll station, and greatly improve the road traffic capacity and vehicle processing speed.

[0111] Referring to Figure 6 ​, after selecting the core factors affecting toll station congestion using the analytic hierarchy process (AHP), the program continues to divide the toll station into detailed regions to ensure that each specific region can be effectively managed according to its characteristics. Then, using the latest YOLOv5 target detection technology, detailed feature data about traffic flow, lane usage, and other key factors are automatically identified and extracted from video surveillance in each region. After integrating this data, the entropy method is used to accurately determine the weights of each key influencing factor. This step is to ensure that the evaluation model can objectively reflect the actual impact of each factor on congestion. Finally, combining these weights and real-time data, a comprehensive toll station congestion index evaluation model is constructed using the TOPSIS method. This model not only accurately assesses the current congestion situation of the toll station, but also predicts future congestion trends, providing scientific data support and decision-making basis for traffic management departments and effectively guiding the implementation of traffic relief measures. The toll gate traffic state recognition method can include the following steps:

[0112] 1. Determine the key influencing factors based on the analytic hierarchy process (AHP) and prepare for further analysis of regional division.

[0113] The analytic hierarchy process (AHP) is a multi-objective decision support tool that combines qualitative and quantitative analysis to handle complex decision-making problems involving multiple factors and multiple objectives. As a weight distribution and priority ranking method, AHP helps decision-makers evaluate the relative importance of various factors by constructing a hierarchical structure to break down the decision-making problem into more manageable parts. It determines the degree of importance between multiple factors based on decision-making experience and gives reasonable weight numbers to calculate the importance order between them.

[0114] (1) Weight calculation: Based on the data and experience of previously implemented toll station-related projects, and combined with expert opinions, the factors affecting toll station congestion are summarized into 12 major categories: toll station vehicle number, toll station lane number, toll station area, toll station lane length, toll station lane width, toll station detection efficiency, toll station traffic efficiency, ETC usage rate, vehicle queue length, weather factors, holiday factors, and equipment failure rate. According to the actual congestion situation of multiple toll stations, Santy's 1-9 scale method is used to compare these 12 factors two by two. First, a 12-level judgment matrix A is constructed. The construction process is as follows: factor i is very important compared to factor j, scoring 5 points, and vice versa, scoring 0.2 points. The product of the two values is equal to 1. Then, the weighted average of the scores is calculated to obtain the judgment matrix A composed of scores.

[0115] Each matrix element represents the importance of factor i relative to factor j, while the diagonal elements of the matrix are 1, i.e. Because any element is equally important compared to itself. If element i is more important than j, then is a value greater than 1 (e.g. 5) and vice versa (e.g. 0.2). Saaty's scale is from 1 to 9, where 1 means two factors are equally important, 3 means slightly important, 5 means obviously important, 7 means strongly important, and 9 means extremely important. Even values between these main values are used to express intermediate feelings. If element j is more important than i, then If the importance score of factor i compared to factor j is 5, then the importance score of j compared to i should be 1 / 5.

[0116] Secondly, find the eigenvector: find the eigenvector by power method and normalize the eigenvector.

[0117] Specifically, select a non-zero initial vector:

[0118]

[0119] Iterative calculation to update the vector , then normalize the vector , and then repeat the above process until converges, that is, when the vectors of two consecutive iterations are very close, stop. Once X converges, the largest eigenvalue .

[0120] Finally, the above process can be analyzed by selecting SPSS tool, that is, input the judgment matrix A into SPSS, according to the above process, the weight value can be calculated, and the analysis result is shown in the following table:

[0121]

[0122] (2) Consistency test, after calculating the weight result by AHP method, get , n is the order of the matrix, that is, CI=0.065. Input the 12-order judgment matrix in the above into the SPSS tool to get the corresponding random consistency RI value of 1.540.

[0123] Calculate the consistency index CR value, . So we get: CR=0.042. The principle of test is that the smaller the CR value, the better the test effect. When the CR value is less than 0.1, it can be judged that this verification meets the consistency requirement and the verification is passed. The following table shows the consistency test result:

[0124]

[0125] ​(3) Result screening: The weight order of the congestion influencing factors at toll stations was calculated by the AHP method as shown in Table 2. Figure 7

[0126] Based on the weight order obtained by the AHP method, the congestion influencing factors at toll stations were clearly classified. The first three factors - the number of vehicles at the toll station, the length of the vehicle queue, and the area of the toll station - were identified as having the most significant direct impact on the congestion state of the toll station, as their total weight accounted for more than 50%. These three key factors directly determine the likelihood and degree of congestion and are the main targets for congestion management and optimization.

[0127] Number of vehicles at the toll station: This factor reflects the traffic volume entering the toll station. High traffic volume can easily cause congestion when the number of lanes or processing capacity is insufficient.

[0128] Length of the vehicle queue: This factor directly reflects the instantaneous processing capacity of the toll station. A long queue usually means low traffic efficiency and is a clear indicator of the degree of congestion.

[0129] Area of the toll station: Adequate area can support more lanes and more reasonable vehicle arrangement, reducing traffic conflicts and congestion.

[0130] The last nine factors, including the number of lanes at the toll station, the width of the toll station lanes, the detection efficiency of the toll station, the traffic efficiency of the toll station, the ETC usage rate, weather factors, holiday factors, and equipment failure rate, while also affecting the congestion state of the toll station, have a relatively small impact compared to the first three main factors. These secondary factors mainly have indirect or auxiliary effects and may only manifest their impact on congestion under certain conditions or in combination with other factors.

[0131] Classifying these factors into primary and secondary factors helps decision-makers and managers focus resources and attention on the most critical factors to develop more effective strategies and measures to reduce congestion. For example, optimizing vehicle number and queue management, expanding or rationally designing the area of the toll station, etc., are direct action directions based on this weight analysis. At the same time, for those factors considered secondary, appropriate optimization measures can be taken as long-term and auxiliary management strategies. This priority classification based on weights ensures the effective use of resources and the scientific nature of congestion management measures, with the ultimate goal of improving the traffic efficiency and vehicle processing capacity of the toll station.

[0132] ​Further analysis of the specific relationships between the factors related to toll station congestion is conducted, and a detailed three-level division of the area in front of the toll station is carried out. This division aims to more finely manage and analyze the traffic flow in each area. To achieve this goal, artificial intelligence technology is used for real-time data analysis, significantly improving the efficiency and accuracy of data processing.

[0133] Real-time video capture: High-definition cameras are used to monitor each entrance and exit area of the toll station around the clock, capturing video images in real time. These video data are the direct source of information for analyzing congestion status, and can capture the real scene of congestion occurrence and the dynamic changes over time.

[0134] Application of YOLOv5 target detection algorithm: By introducing advanced YOLOv5 target detection technology, the system can quickly and accurately identify and distinguish key feature data from video images. This includes the number of vehicles, queue length, lane occupancy, and possible abnormal traffic behavior such as illegal lane changing or sudden stopping. The application of YOLOv5 algorithm is particularly suitable for complex traffic scene analysis due to its high efficiency and powerful real-time processing capability.

[0135] Data analysis and decision support: The collected data will be transmitted in real time to the central processing system, where AI algorithms will conduct in-depth analysis of the data to extract congestion patterns and identify congestion trends. During this process, AI not only processes image data, but also combines various sensor information and historical traffic data to provide more comprehensive traffic flow analysis.

[0136] 2. Division of toll station area

[0137] Based on the three key influencing factors determined by the analytic hierarchy process (AHP): toll station vehicle number, vehicle queue length, and toll station area, we divide the toll station into three specific areas for more targeted management and optimization. These areas are:

[0138] A lane area: This area is mainly composed of ETC / MTC toll islands, barrier equipment, lane markings, etc. It is the core operation area of the toll station, directly affecting the efficiency and vehicle handling capacity of the toll station. The boundary of A area is determined by the presence or absence of lane markings, which distinguishes between A and B areas, ensuring the normal operation and safety management of each lane.

[0139] B buffer area: Located between the A lane area and the C ramp area, this buffer area usually has no lane markings, and its main function is to provide a buffer space for vehicles entering the toll station to reduce the pressure and potential congestion caused by direct entry into the toll island. This area is also a sorting area for vehicles waiting to enter the toll island, helping to orderly queue and distribute vehicles.

[0140] C Loop Area: This area is the road where vehicles enter the toll station from the main highway, serving as the key passage connecting the highway and the toll station. In this area, vehicles typically need to slow down to 30-40 km / h to meet the safety requirements of the toll station. The design and management of the C area directly affect the smoothness and safety of vehicle entry into the toll station.

[0141] 3. Feature Extraction Based on YOLOv5 Algorithm

[0142] YOLOv5 is an advanced single-stage object detection algorithm that is highly praised for its high flexibility, fast recognition ability, and simple deployment process. The core advantages of this algorithm include adaptive anchor box calculation, adaptive image scaling, and providing four different network structure sizes (small, medium, large, xlarge) to adapt to different operation and precision needs.

[0143] Application Strategy for Toll Station:

[0144] Region Division and ROI Setting: Based on the previously defined A lane area, B buffer area, and C loop area, we frame the ROI (Region of Interest) in each area, which is considered to be the most frequent or most likely place for vehicle activity or congestion.

[0145] Feature Extraction and Learning: Using the efficient algorithm of YOLOv5, real-time analysis of video pictures in each ROI is performed to automatically extract key features such as vehicle number, vehicle type, and vehicle speed. This step is completed through the algorithm's adaptive anchor box technology and deep learning model, ensuring the accuracy and real-time nature of the detection.

[0146] Building Feature Library and Training Model: The structural features and feature data extracted from the video are used to build a feature library of vehicles at the toll station. This database is used to train the vehicle detection model, enabling it to identify and classify different types of vehicles.

[0147] Real-time Video Detection and Analysis Results: Based on the trained vehicle detection model, the vehicle flow status of each area can be monitored and analyzed in real time, with the model's required data being fed back in a timely manner. The model's prediction results can be provided to traffic managers through a comprehensive display interface.

[0148] (1) A Lane Area

[0149] First, we capture the ROI (Region of Interest) areas of each lane using high-precision camera equipment, ensuring that the visual range of each lane is accurately captured. Using the already trained and optimized YOLOv5 object detection algorithm model, the system can obtain the coordinate values of each vehicle and its entry number in real time. These data are used for further analysis and comparison of vehicle coordinates with the preset lane coordinate range, ensuring that each vehicle is correctly classified into the corresponding lane.

[0150] Vehicle statistics and queue length estimation process:

[0151] Vehicle identification and classification: The vehicle coordinates and type (small or large vehicle) information detected by the YOLOv5 algorithm are used for classification and statistics. This step is crucial in distinguishing different types of vehicles to accurately calculate their impact on traffic flow.

[0152] Lane vehicle statistics: Based on the comparison of vehicle coordinates with lane coordinate ranges, the system counts the number of vehicles per lane. This statistical method allows the traffic management system to monitor the number of vehicles in each lane in real time, assessing traffic flow and potential congestion points.

[0153] Queue length estimation: Depending on the vehicle type, the system estimates the queue length of each lane. Small vehicles are typically 1.5 meters long, while large vehicles are 9 meters long on average, with an estimated vehicle spacing of 0.5 meters. Based on the type and number of vehicles in a lane, the total queue length of each lane is calculated. For example, if there are 5 small vehicles and 2 large vehicles in a lane, the queue length calculation is:

[0154]

[0155] Data aggregation and feedback: Total number of vehicles in the area, number of vehicles per enabled lane, and queue length per enabled lane.

[0156] (2) B Buffer Area

[0157] First, we ensure that the area of each labeled region of interest (ROI) is accurately captured by high-precision monitoring equipment, and record the actual area values in these regions. Then, using the already trained and optimized YOLOv5 object detection algorithm model, the system can obtain the coordinate values of each vehicle and its entry number in real time. This information will be used to compare vehicle coordinates with the coordinate range of each region, accurately counting the number of vehicles per region.

[0158] Data processing and vehicle statistics process:

[0159] Vehicle identification and coordinate acquisition: Through the YOLOv5 algorithm, the system identifies the coordinate position and entry number of each vehicle from video monitoring, which is the key data for determining vehicle location and region matching.

[0160] Area and vehicle comparison: Compare the coordinates of the vehicles with the coordinate range of the labeled ROI areas to ensure that each vehicle is correctly categorized into the corresponding area. This step is completed through high-precision spatial analysis to ensure data accuracy.

[0161] Statistical vehicle number by area: The number of vehicles in each ROI area is counted based on the coordinate comparison results, which accurately calculates the vehicle density and distribution in each area.

[0162] Data summary and feedback: Total buffer area and total number of vehicles in the buffer area.

[0163] (3) C ramp area

[0164] First, the labeled ROI (Region of Interest) area data in the ramp section is obtained through high-precision camera equipment to ensure accurate monitoring of each key area. Using the trained and optimized YOLOv5 target detection algorithm model, the system can capture the coordinate value of each vehicle and its entry number in real time, and record the vehicle license plate information and entry time.

[0165] Vehicle stay time analysis and vehicle statistics process:

[0166] License plate information and entry time matching: Compare the license plate information and entry time detected and recorded by the YOLOv5 algorithm with historical data to determine whether the vehicle has stayed in the ramp area for more than the specified time. This step is crucial in identifying and handling abnormal stay behaviors that may cause congestion.

[0167] Screening compliant vehicle information: Screen out vehicle information that does not exceed the stay time from the data. These vehicles are considered to be in normal flow and will not adversely affect traffic flow.

[0168] Statistical vehicle number by area: Compare the vehicle coordinates with the area coordinate range of the ramp section to count the number of vehicles in each area. This statistic helps managers understand the vehicle distribution and density in each area, which is crucial for adjusting traffic control measures.

[0169] Data summary and feedback: Total ramp area and total number of vehicles in the ramp area.

[0170] 4. Determine the index weight of the key influencing factors using the entropy method

[0171] The entropy method obtains dynamic accumulation data of 12 influencing factors such as historical sub-period vehicle number, area, and queue length of toll stations to judge the dispersion degree of each index. The greater the dispersion degree, the greater the influence of the index on comprehensive evaluation. The use of the entropy method to determine the index weight is divided into the following steps.

[0172] (1) Each index is normalized by maximum and minimum normalization, and adjusted considering the positive and negative directions of the index;

[0173] The specific formula used is:

[0174] wherein is the original value of the ith sample at the jth index, is the minimum value and maximum value of the jth index, respectively.

[0175] (2) After standardization, calculate the proportion of the ith sample value under the jth index to the index:

[0176]

[0177] wherein n is the number of samples, is the standardized value of the ith sample at the jth index.

[0178] (3) Calculate the entropy value of each index using the proportion data. The greater the entropy value, the more uniform the data of the index, and the less important the index is to the comprehensive evaluation. Calculate the entropy value of the jth index:

[0179]

[0180] wherein k is a constant, .

[0181] (4) Calculate the difference coefficient:

[0182]

[0183] (5) Calculate the weight of each index. The greater the weight, the greater the variability of the index, and the more important the index is to the comprehensive evaluation:

[0184]

[0185] wherein m is the total number of indexes, and the number of indexes selected this time is 12.

[0186] 5. Construct an evaluation model using the TOPSIS method

[0187] The TOPSIS method can fully utilize the information of each index data and accurately reflect the gap between each evaluation scheme. The use of this method is shown as follows.

[0188] (1) Using historical data, combined with the toll station traffic management objectives for analysis, such as the length of the vehicle queue, considering the toll station queue more than 200 meters will be forced to free open traffic Department of management requirements, in order to ensure that the toll and queue controllable premise, determine the optimal TOPSIS as the best and worst solution; for example, the number of lanes when the toll station is less than 3, it is easy to cause congestion; toll station vehicle number more than 6, it is easy to cause congestion; toll station area if less than 5000 square meters of the peak period is prone to congestion, etc.

[0189] (2) using the optimal and worst value of the fixed data for normalization processing, and considering the positive and negative adjustment of the index:

[0190]

[0191] Wherein is the value of the normalized impact factor.

[0192] (3) using the Euclidean distance calculation and optimal worst target distance, and multiplied by the weight:

[0193]

[0194] And is the ideal and non-ideal value of the jth index, is the weight of the jth index, is the normalized value of the ith scheme in the jth index.

[0195] (4) calculate the proximity of each evaluation object and the optimal scheme:

[0196]

[0197] (5) based on the proximity value of 12 evaluation objects, calculate the weighted average proximity value C

[0198]

[0199] Wherein is the proximity value of the ith impact factor, is the weight value determined according to the entropy method before.

[0200] 5, data index and score evaluation

[0201] The YOLOv5 target detection algorithm is used to model and train the video data of three regions of the toll station, and the specified feature data is extracted. Then, the congestion index evaluation model of the toll station is constructed, and the final result is calculated. The congestion index of the toll station under the condition of no car is defined as 0, and the congestion index of the toll station under the condition of full car in the ramp area is defined as 1.0. The corresponding congestion degree of the toll station congestion index score is shown in the following table:

[0202]

[0203] Based on the above threshold standard, the model calculation can be performed on the toll stations of different sections to obtain the busy index state within a day, within a week, and during holidays, and the results can be presented and fed back in a visual graphical manner. Referring to Figure 8 、 Figure 9 、 Figure 10 and Figure 11 , the results of the data analysis and verification of the idle period (5:00-8:00), the afternoon period on weekdays (15:00-18:00), the afternoon period on weekends (15:00-18:00), and the afternoon period on holidays (15:00-18:00) of one of the highway exit toll stations can be obtained.

[0204] As can be seen from the above schematic diagram, the congestion index results of the toll station in the idle period are all below the basic smooth line; the congestion index results of the toll station in the afternoon period on weekdays are mostly below the moderate congestion line and above the light congestion line, and the results of a few time points are above the moderate congestion line; the congestion index results of the toll station in the afternoon period on weekends are mostly above the moderate congestion line and below the severe congestion line, and a small part of the results are above the severe congestion line; the congestion index results of the toll station in the afternoon period on holidays are all above the moderate congestion line and below the severe congestion line.

[0205] The above calculation is compared with the actual situation, and it is found that the estimated situation calculated by the method provided by the embodiment of the present application is accurate and can be implemented.

[0206] 6. Provide a solution plan suggestion according to the data index

[0207] The results analyzed by the above model are displayed and prompted on the screen of the operation and maintenance personnel, including not only the congestion prediction graph of different time periods, but also the close data of each evaluation object and the optimal scheme, which are used to give a solution plan for reference.

[0208] When the close data of the influencing factor lane number is the worst, the lane configuration can be optimized: the number of open lanes or lane use strategy is adjusted according to the traffic flow;

[0209] When the proximity data of the influencing factor ETC usage rate is the worst, consider improving ETC usage rate: increase ETC usage rate through policy incentives or user education, reduce manual toll congestion;

[0210] When the proximity data of the influencing factor vehicle queue length is the worst, consider improving equipment and technology: update obsolete equipment and improve the efficiency of detection and charging technology.

[0211] At the same time, we can also consider improving the traffic guidance system: use traffic guidance systems near toll stations to reasonably allocate vehicle flow to different lanes.

[0212] After implementing the improvement measures, continue to monitor the performance of each toll station, evaluate the effectiveness of the management measures, and adjust and optimize according to the actual situation. Through continuous performance monitoring and feedback, we can ensure that the congestion management strategy of the toll station is always effective.

[0213] Develop emergency opening strategies for unpredictable peak traffic or special events (such as accidents or extreme weather) to quickly relieve sudden congestion.

[0214] Embodiment six

[0215] Figure 12 is a structural schematic diagram of a gate traffic state recognition device provided by an embodiment of the present application, as Figure 12 shown, the device comprises:

[0216] The core factor module 510 is used to obtain the traffic influencing factors of the target gate and identify the core influencing factors in the traffic influencing factors.

[0217] The image acquisition module 520 is used to determine the data acquisition area of the target gate based on the core influencing factors and acquire the traffic image information in the data acquisition area.

[0218] The feature extraction module 530 is used to extract the traffic state features of each traffic image information according to the preset feature extraction rule.

[0219] The state recognition module 540 is used to generate the traffic congestion recognition result of the target gate according to the traffic state features of different data acquisition areas.

[0220] The embodiment of the present application acquires traffic influence factors of a target gate through a core factor module, and determines core influence factors in the traffic influence factors. An image acquisition module determines a data acquisition area of the target gate based on the core influence factors, and acquires traffic image information in the data acquisition area. A feature extraction module extracts traffic state features in the traffic image information. A state recognition module determines a traffic congestion recognition result of the target gate according to the traffic state features of different data acquisition areas. The embodiment of the present application can improve the accuracy of traffic state recognition by quantitatively analyzing traffic influence factors and determining core influence factors affecting gate traffic. The traffic state features of the data acquisition area are used to predict and manage traffic flow, which facilitates optimization of road and toll station use efficiency, reduces congestion and waiting time, improves smoothness of traffic, and improves user experience.

[0221] In some embodiments of the application, the core factor module 510 includes:

[0222] A data acquisition unit is configured to acquire gate traffic data of at least one data source according to the target gate.

[0223] A factor division unit is configured to classify and divide the gate traffic data according to a preset factor into the traffic influence factors.

[0224] A core recognition unit is configured to call an analytic hierarchy process rule to determine a weight value of each traffic influence factor, and select a threshold number of traffic influence factors as the core influence factors according to the weight value.

[0225] In some embodiments of the application, a consistency verification module is further included, configured to acquire a consistency verification result of the weight value of each traffic influence factor, and determine that the consistency verification result is passed.

[0226] In some embodiments of the application, the image acquisition module 520 includes:

[0227] A region division unit is configured to acquire a three-dimensional space model of a space where the target gate is located, and determine the data acquisition area corresponding to the core influence factors in the three-dimensional space model.

[0228] An information acquisition unit is configured to call a preset image acquisition device to acquire the traffic image information in each data acquisition area.

[0229] In some embodiments of the application, the feature extraction module 530 is configured to perform at least one of the following:

[0230] According to a target detection algorithm, the number of total vehicles, the number of vehicles in an active lane, and the queue length in the active lane in the traffic image information are recognized as the traffic state features.

[0231] identify the buffer area and the total number of vehicles in the buffer area in the traffic image information as the traffic state features according to a target detection algorithm;

[0232] identify the total ramp area and the total number of parked vehicles in the ramp in the traffic image information as the traffic state features according to a target detection algorithm.

[0233] In some embodiments of the application, the state recognition module 540 comprises:

[0234] a model extraction unit configured to obtain a traffic index evaluation model and index weights of historical traffic influence factors, wherein the traffic index evaluation model and the index weights are generated based on historical traffic data.

[0235] a feature normalization unit configured to determine optimal index values and worst index values of the traffic state features in the traffic index evaluation model according to the traffic influence factors of the traffic state features, and normalize the traffic state features according to the optimal index values and the worst index values.

[0236] a distance determination unit configured to determine optimal Euclidean distances and worst Euclidean distances between the traffic state features and the optimal index values and the worst index values, respectively.

[0237] a closeness value unit configured to determine a sum of the optimal Euclidean distances and the worst Euclidean distances, and determine a ratio of the worst Euclidean distances to the sum as a closeness value of the traffic state features.

[0238] a result obtaining unit configured to determine a weighted average closeness value based on the closeness values of the traffic state features and the index weights, and determine a congestion condition of the weighted average closeness value as a traffic congestion recognition result according to a preset congestion degree table.

[0239] On the basis of the above-mentioned embodiments of the application, the model extraction unit obtains the traffic index evaluation model and the index weights of the historical traffic influence factors, comprising:

[0240] obtain the historical traffic data, and classify and divide the historical traffic data into different historical traffic influence factors according to preset factors;

[0241] standardize the historical traffic influence factors based on maximum values and minimum values in the historical traffic influence factors for each historical traffic influence factor;

[0242] determine entropy values of each historical traffic influence factor, and obtain difference coefficients corresponding to the entropy values;

[0243] determine the index weights of the historical traffic influence factors as ratios of each difference coefficient to all difference coefficients;

[0244] The maximum value and the minimum value of each of the historical traffic influencing factors are saved as the traffic index evaluation model.

[0245] The gate traffic state recognition device provided by the embodiments of the present application can execute the gate traffic state recognition method provided by any of the embodiments of the present application, and has the corresponding function modules and beneficial effects of the execution method.

[0246] Embodiment Seven

[0247] Figure 13 is a structural schematic diagram of an electronic device for implementing the gate traffic state recognition method of the embodiments of the present application. The electronic device is intended to represent various forms of digital computers, such as laptops, desktops, tablets, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular telephones, smart phones, wearable devices (such as helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be examples only, and are not intended to limit the implementations of the present application described and / or claimed in this document.

[0248] As shown in Figure 13 , the electronic device 10 includes at least one processor 11, and a memory, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., which is communicatively connected to the at least one processor 11. The memory stores a computer program that can be executed by the at least one processor. The processor 11 can execute various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or loaded from the storage unit 18 into the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0249] A plurality of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, a loudspeaker, etc.; a storage unit 18, such as a magnetic disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network, such as the Internet, and / or various telecommunication networks.

[0250] The processor 11 can be various general and / or special purpose processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various specialized artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The processor 11 performs various methods and processes described above, such as the gate traffic state recognition method.

[0251] In some embodiments, the gate traffic state recognition method can be implemented as a computer program tangibly embodied in a computer readable storage medium, such as the storage unit 18.

[0252] In some embodiments, part or all of the computer program can be loaded and / or installed onto the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded onto the RAM 13 and executed by the processor 11, one or more steps of the gate traffic state recognition method described above can be performed. Alternatively, in other embodiments, the processor 11 can be configured to perform the gate traffic state recognition method by any other suitable means, such as by means of firmware.

[0253] Various implementations of the systems and techniques described above can be realized in digital electronic circuitry, integrated circuitry, a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), a system on a chip (SOC), a programmable logic device (CPLD), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include implementation in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device.

[0254] Computer programs used to implement the gate traffic state recognition method of the present application can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the computer program, when executed, implements the functions / operations specified in the flowcharts and / or the block diagrams. The computer program can be executed entirely on a machine, partially on a machine, partially on a machine as a stand-alone software package, and partially on a remote machine or a server.

[0255] In the context of the present application, a computer-readable storage medium can be a tangible medium that can contain or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. A computer-readable storage medium can include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium can be a machine-readable signal medium. More specific examples of a machine-readable storage medium will include one or more lines of a program of instructions in a transitory signal, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0256] To provide for interaction with a user, the systems and techniques described here can be implemented on an electronic device having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the electronic device. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form, including acoustic, speech, or tactile input.

[0257] The systems and techniques described here can be implemented in a computing system that includes a back end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front end component (e.g., a user computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back end, middleware, or front end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), blockchain network, and the Internet.

[0258] The computing system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system, to solve the defects of large management difficulty and weak business scalability in traditional physical host and VPS service.

[0259] It should be understood that the various forms of flow shown above can be used to reorder, add or delete steps. For example, each step described in the present application can be executed in parallel, sequentially or in a different order, as long as the desired results of the technical solutions of the present application can be achieved, which is not limited herein.

[0260] The above detailed description does not constitute a limitation on the scope of protection of the present application. Those skilled in the art should understand that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent replacements and improvements made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A method of identifying a traffic state at a gate, characterized by, The method comprises: acquiring traffic influencing factors of a target gate, and identifying core influencing factors within the traffic influencing factors; determining a data collection area of the target gate based on the core influencing factors, and acquiring traffic image information within the data collection area; extracting traffic state features of each of the traffic image information according to preset feature extraction rules; generating a traffic congestion identification result of the target gate according to the traffic state features of different data collection areas.

2. The method of claim 1, wherein, The acquisition of the traffic influencing factors of the target gate and the identification of the core influencing factors within the traffic influencing factors comprise: extracting gate traffic data of at least one data source according to the target gate; classifying and dividing the gate traffic data into the traffic influencing factors according to preset factors; determining weight values of each of the traffic influencing factors by calling an analytic hierarchy rule, and selecting a threshold number of the traffic influencing factors as the core influencing factors according to the weight values.

3. The method of claim 1 or 2, wherein, Further comprising: acquiring a consistency test result of the weight values of each of the traffic influencing factors, and determining that the consistency test result is passed.

4. The method of claim 1, wherein, The determination of the data collection area of the target gate based on the core influencing factors and the acquisition of the traffic image information within the data collection area comprise: acquiring a three-dimensional space model of a space where the target gate is located, and determining the data collection area corresponding to the core influencing factors within the three-dimensional space model; collecting the traffic image information within each of the data collection areas by calling a preset image collection device.

5. The method of claim 1, wherein, The extraction of the traffic state features of each of the traffic image information according to the preset feature extraction rules comprises at least one of the following: identifying the total number of vehicles in a region, the number of vehicles in an active lane, and the queue length in the active lane within the traffic image information as the traffic state features according to a target detection algorithm; identifying the buffer area and the total number of vehicles in the buffer area within the traffic image information as the traffic state features according to a target detection algorithm; identifying the total area of a ramp and the total number of parked vehicles within the traffic image information as the traffic state features according to a target detection algorithm.

6. The method of claim 1, wherein, The generation of the traffic congestion identification result of the target gate according to the traffic state features of different data collection areas comprises: acquiring a traffic index evaluation model and an index weight of each historical traffic influencing factor, wherein the traffic index evaluation model and the index weight are generated based on historical traffic data; determining optimal and worst index values in the traffic index evaluation model according to the traffic influencing factors of the traffic state features, and normalizing the traffic state features according to the optimal and worst indexes; determining optimal and worst Euclidean distances between the traffic state features and the optimal and worst indexes, respectively; determining a distance sum of the optimal and worst Euclidean distances, and taking a ratio of the worst Euclidean distance to the distance sum as a closeness value of the traffic state features; determining a weighted average closeness value based on the closeness values of each of the traffic state features and the index weight; According to the preset congestion degree table, a congestion situation of the weighted average approximation value is searched as the traffic congestion recognition result.

7. The method of claim 6, wherein, The traffic index evaluation model and the index weight of each historical traffic influence factor are obtained, including: The historical traffic data is obtained, and the historical traffic data is classified and divided into different historical traffic influence factors according to a preset factor; For each historical traffic influence factor, the historical traffic influence factor is standardized based on the maximum value and the minimum value in the historical traffic influence factor; The entropy value of each historical traffic influence factor is determined, and a difference coefficient corresponding to the entropy value is obtained; The ratio of each difference coefficient to all difference coefficients is taken as the index weight of the historical traffic influence factor; The maximum value and the minimum value of each historical traffic influence factor are saved as the traffic index evaluation model.

8. A gateway traffic state recognition device characterized by comprising: The device comprises: A core factor module is configured to obtain traffic influence factors of a target gate and identify core influence factors in the traffic influence factors; An image acquisition module is configured to determine a data acquisition area of the target gate based on the core influence factors and obtain traffic image information in the data acquisition area; A feature extraction module is configured to extract traffic state features of each traffic image information according to a preset feature extraction rule; A state recognition module is configured to generate a traffic congestion recognition result of the target gate according to the traffic state features of different data acquisition areas.

9. An electronic device, comprising: The electronic device comprises: At least one processor; and A memory connected in communication with the at least one processor; wherein The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to execute the gate traffic state recognition method in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer instructions for enabling the processor to implement the gate traffic state recognition method in any one of claims 1-7 when executed.

Images