Intelligent tracking method and system for vehicle trajectory

Abstract

The application relates to the technical field of intelligent identification, and provides an intelligent tracking method and system for a vehicle track, which comprises the following steps: vehicle state monitoring is performed through a traffic monitoring system, vehicle features in a monitoring picture are traversed according to vehicle tracking features, and a monitoring target vehicle is determined; vehicle information, driving position information and driving speed information of the monitoring target vehicle are collected; vehicle driving performance evaluation is performed, and vehicle driving parameters are determined; driving state prediction of the monitoring target vehicle is performed according to the driving position information, the driving speed information and the vehicle driving parameters, vehicle prediction monitoring tracks are determined; monitoring information of the monitoring target vehicle is continuously collected, track deviation degree calculation is performed based on the vehicle prediction monitoring tracks, tracking track correction is performed based on the track deviation degree, and vehicle tracking information is obtained. The application can solve the technical problem that the accuracy of vehicle tracking according to a vehicle track is low in the prior art, and can improve the accuracy of vehicle tracking.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of collaborative office technology, in particular to a vehicle trajectory intelligent tracking method and system. BACKGROUND

[0002] With the acceleration of urbanization, the road traffic volume increases year by year, and traffic congestion has become a common problem in urban traffic. In order to solve this problem, various intelligent traffic systems are continuously developed and researched, and vehicle tracking and behavior recognition technology is an important part of intelligent traffic system. The existing vehicle tracking method according to vehicle trajectory has low vehicle feature comparison precision, and the vehicle trajectory calculation method is imperfect, resulting in low vehicle tracking accuracy.

[0003] In summary, the prior art has the technical problem of low vehicle tracking accuracy according to vehicle trajectory. SUMMARY

[0004] Therefore, it is necessary to provide a vehicle trajectory intelligent tracking method and system to solve the above technical problems.

[0005] A vehicle trajectory intelligent tracking method, comprising: vehicle state monitoring by a traffic monitoring system, vehicle feature traversal of a monitoring picture according to vehicle tracking features, and determination of a monitoring target vehicle; vehicle information, driving position information, and driving speed information of the monitoring target vehicle are collected; vehicle driving performance evaluation is performed according to the vehicle information to determine vehicle driving parameters; driving state prediction of the monitoring target vehicle is performed according to the driving position information, driving speed information, and vehicle driving parameters to determine vehicle prediction monitoring trajectory; monitoring information of the monitoring target vehicle is continuously collected, and trajectory deviation calculation is performed based on the vehicle prediction monitoring trajectory, tracking trajectory correction is performed based on the trajectory deviation, and vehicle tracking information is obtained.

[0006] In one embodiment, it further includes setting multiple types of tracking features, collecting samples of each type based on the multiple types of tracking features to obtain a multiple-type sample set, extracting features from the multiple-type sample set to obtain a feature set of each sample, calculating feature weights of the feature set of each sample, setting a sample feature traversal sequence based on the feature weights, and traversing and comparing the sample feature traversal sequence as vehicle tracking features.

[0007] In one embodiment, further comprising: performing feature recognition degree analysis to determine feature saliency; performing feature repetition rate calculation to determine feature repetition rate; performing tracking rule analysis based on the multi-type tracking features to determine rule constraint features; performing feature evaluation on the feature saliency, feature repetition rate, and rule constraint features according to a preset weight setting value to obtain a feature evaluation value; and taking the feature evaluation value as a weight coefficient to obtain a weight value of each feature.

[0008] In one embodiment, further comprising: performing position area road extraction through an electronic map system according to the driving position information to obtain driving road information including road condition information and passing road nodes; performing driving strategy variation probability prediction to obtain driving direction prediction information according to the driving speed information and vehicle driving parameters; and performing driving trajectory prediction according to the driving road information and the driving direction prediction information.

[0009] In one embodiment, further comprising: determining a targeting point and a targeting distance based on the vehicle predicted monitoring trajectory and the monitoring information; performing trajectory tracking based on the targeting point and the targeting distance through a trajectory tracking algorithm module to obtain a vehicle trajectory calculation result; and obtaining a trajectory deviation degree according to the vehicle trajectory calculation result and the vehicle predicted monitoring trajectory.

[0010] In one embodiment, further comprising: extracting a first target image of the monitoring information of the target vehicle; reading a pre-targeting pixel number based on the targeting point, and determining a center of a longest diameter pixel number as a circle center according to the targeting distance and the target vehicle, and drawing a target circle with the pre-targeting pixel number as a radius; taking an intersection of the target circle and the vehicle predicted monitoring trajectory as a pre-targeting pixel point; obtaining a first actual motion direction of a pixel area, and taking an included angle between the first actual motion direction and the pre-targeting pixel point as a deviation angle; calculating a first actual turning angle based on the longest diameter pixel number, the pre-targeting pixel number, and the deviation angle; extracting a second target image of the monitoring information of the target vehicle, and analyzing the second target image to obtain a second actual turning angle, wherein the first target image and the second target image are adjacent images; and obtaining the vehicle trajectory calculation result based on the first actual turning angle and the second actual turning angle.

[0011] An intelligent tracking system of a vehicle trajectory, comprising:

[0012] A monitoring target vehicle determination module, configured to perform vehicle state monitoring through a traffic monitoring system, perform monitoring picture vehicle feature traversal according to vehicle tracking features, and determine a monitoring target vehicle.

[0013] An information collection module, configured to collect vehicle information, driving position information, and driving speed information of the monitoring target vehicle.

[0014] a vehicle driving parameter determination module configured to determine a vehicle driving parameter according to the vehicle information;

[0015] a vehicle predicted monitoring trajectory determination module configured to determine a vehicle predicted monitoring trajectory according to the driving position information, the driving speed information and the vehicle driving parameter;

[0016] a vehicle tracking information obtaining module configured to continuously collect monitoring information of the monitoring target vehicle, and perform trajectory deviation calculation based on the vehicle predicted monitoring trajectory, perform tracking trajectory correction based on the trajectory deviation, and obtain vehicle tracking information.

[0017] The intelligent vehicle tracking method and system can solve the technical problem of low accuracy of vehicle tracking according to vehicle trajectory in the prior art. First, the vehicle state is monitored by the traffic monitoring system, the vehicle features of the monitoring screen are traversed according to the vehicle tracking features, and the monitoring target vehicle is determined. The vehicle information, driving position information and driving speed information of the monitoring target vehicle are collected. The vehicle driving parameter is determined according to the vehicle information. The driving state of the monitoring target vehicle is predicted according to the driving position information, the driving speed information and the vehicle driving parameter, and the vehicle predicted monitoring trajectory is determined. The monitoring information of the monitoring target vehicle is continuously collected, and the trajectory deviation calculation is performed based on the vehicle predicted monitoring trajectory. The tracking trajectory is corrected based on the trajectory deviation, and the vehicle tracking information is obtained. The accuracy of vehicle tracking according to vehicle trajectory can be improved by the above method.

[0018] The above description is only a summary of the technical solutions of the present application. In order to more clearly understand the technical means of the present application, the above description can be implemented according to the content of the specification, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the following specific embodiments of the present application are described. BRIEF DESCRIPTION OF DRAWINGS

[0019] Figure 1 A flowchart of an intelligent vehicle tracking method is provided for the present application;

[0020] Figure 2 A flowchart of determining a monitoring target vehicle in an intelligent vehicle tracking method is provided for the present application;

[0021] Figure 3 A structural diagram of an intelligent vehicle tracking system is provided for the present application.

[0022] Explanation of reference signs: monitoring target vehicle determination module a, information collection module b, vehicle driving parameter determination module c, vehicle predicted monitoring trajectory determination module d, vehicle tracking information obtaining module e. DETAILED DESCRIPTION

[0023] In order to make the purpose, technical scheme and advantages of the present application clearer, the present application will be further described in detail below in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application and do not limit the present application.

[0024] As shown in Figure 1 , the present application provides an intelligent tracking method of vehicle trajectory, comprising:

[0025] The vehicle state is monitored by the traffic monitoring system, the vehicle feature traversal of the monitoring picture is performed according to the vehicle tracking feature, and the monitoring target vehicle is determined;

[0026] As shown in Figure 2 , in one embodiment, it further comprises:

[0027] A plurality of types of tracking features are set, and a plurality of types of sample sets are obtained based on the plurality of types of tracking features;

[0028] The feature extraction is performed on the plurality of types of sample sets to obtain the feature set of each sample;

[0029] The vehicle trajectory refers to the position information recorded by the vehicle within a period of time, which is usually obtained through the global positioning system (GPS) or other positioning technology. According to these trajectory data, functions such as real-time monitoring of vehicles, analysis of driving routes, and identification of vehicle behavior can be realized. The method provided by the present application is used for vehicle tracking according to vehicle trajectory, which is used to improve the accuracy of intelligent tracking of vehicles, and the method provided by the present application is embodied in an intelligent tracking system of vehicle trajectory.

[0030] First, connect the traffic monitoring system of the target area, the target area refers to the driving area of the target monitoring vehicle, and the traffic monitoring system refers to a system for monitoring and managing traffic flow, traffic state and traffic safety. This kind of system usually uses sensors, cameras, GPS devices and other technologies to collect traffic-related data, and then analyzes and processes these data to provide real-time traffic information, traffic flow prediction, traffic condition evaluation and traffic event identification functions. The vehicle state is monitored by the traffic monitoring system.

[0031] Set multi-type vehicle tracking features, wherein the vehicle tracking features refer to key attributes and indicators for identifying and tracking vehicles. These features are usually extracted from vehicle trajectory data to describe the motion and behavior of vehicles. The multi-type vehicle tracking features include vehicle historical trajectory, vehicle outer dimensions, vehicle position, etc. And according to the multi-type tracking features, sample collection of each type of vehicle features in the monitoring picture is carried out to obtain a multi-type sample set. Then feature extraction is performed on the multi-type sample set, which refers to classifying the multi-type sample set according to the sample feature type to obtain a feature set of each sample.

[0032] Feature weight calculation is performed on the feature set of each sample respectively, and a sample feature traversal sequence is set based on the feature weight;

[0033] In one embodiment, it further comprises:

[0034] Conducting recognition degree analysis on the features to determine feature salience;

[0035] Conducting repetition rate calculation on the features to determine feature repetition rate;

[0036] Conducting tracking rule analysis based on the multi-type tracking features to determine rule constraint features;

[0037] Conducting feature evaluation on the feature salience, feature repetition rate, and rule constraint features according to the preset weight setting value to obtain a feature evaluation value;

[0038] Taking the feature evaluation value as a weight coefficient to obtain the weight value of each feature.

[0039] Feature weight calculation is performed on the feature set of each sample respectively. First, recognition degree analysis is performed on the features, which refers to judging the feature data quality to determine the feature salience, wherein the higher the feature data quality, the higher the feature salience. Then, repetition rate calculation is performed on the features, which refers to measuring the proportion of repeated features in a group of features to determine the feature repetition rate, wherein the more times the feature appears, the higher the feature repetition rate. Tracking rule analysis is performed according to the multi-type tracking features, which refers to analyzing and judging the vehicle tracking rules, wherein the vehicle tracking rules can be defined according to actual needs and scenarios, for example: overspeed rule: if the speed of a vehicle exceeds the set speed limit value, start tracking the vehicle and record its position and speed information, which can be used to monitor traffic violations such as overspeed driving; abnormal driving rule: if a vehicle exhibits abnormal driving behavior such as frequent lane changing, sudden braking, etc., start tracking the vehicle and record the abnormal driving information. This helps to identify traffic accident risks or unsafe driving behaviors. Rule constraint features are determined, which refer to the limiting features in the vehicle tracking rules, such as overspeed driving, abnormal lane changing, etc.

[0040] Then the feature evaluation is performed on the feature distinctness, feature repetition rate and rule constraint feature according to preset weight setting values, the preset weight setting values can be set by the person skilled in the art based on the influence degree of the feature on vehicle tracking, wherein the greater the influence degree of the feature on the vehicle, the greater the feature evaluation value, the feature evaluation value can be calculated by the existing coefficient of variation method, the coefficient of variation method (CV) is a statistical method for comparing the variability of different data sets. It is the ratio of the standard deviation of the data to the mean, used to measure the dispersion of the data relative to its mean. The coefficient of variation method is a commonly used weight setting method for the person skilled in the art, which is not described here. Obtain the feature evaluation value. And the feature evaluation value is used as a weight coefficient to obtain the weight of each feature. And set the sample feature traversal sequence according to the feature weight, wherein the greater the weight of the feature, the earlier the position of the feature in the sample feature traversal sequence.

[0041] The sample feature traversal sequence is traversed and compared as the vehicle tracking feature.

[0042] The sample feature traversal sequence is traversed and compared as the vehicle tracking feature, and the monitoring picture vehicle feature is traversed and compared to obtain a preset feature comparison threshold, the preset feature comparison threshold can be set by the actual situation, the monitoring vehicle in the traversal comparison result that meets the preset feature comparison threshold is taken as the monitoring target vehicle, and the monitoring target vehicle is obtained. Through the monitoring target vehicle, the basis for the next step of vehicle intelligent tracking is provided.

[0043] Collecting vehicle information, driving position information and driving speed information of the monitoring target vehicle;

[0044] Performing vehicle driving performance evaluation according to the vehicle information to determine vehicle driving parameters;

[0045] The vehicle information, driving position information and driving speed information of the monitoring target vehicle are obtained through a plurality of sensors, cameras and other devices in the traffic monitoring system; the vehicle information includes license plate number, vehicle type, vehicle brand and model, vehicle color, vehicle size and other information, the driving position information can be obtained through GPS information, the global positioning system (GPS) is the most commonly used technology for obtaining vehicle position information. Modern vehicles are usually equipped with GPS devices, which can accurately obtain the longitude and latitude coordinates of the vehicle, thereby determining the position of the vehicle; the vehicle driving speed information can be obtained by measuring through radar, camera and other monitoring devices.

[0046] Then according to the vehicle information, the vehicle driving performance evaluation is performed, the vehicle driving parameter is determined, the vehicle information is the description of the basic information of the vehicle, including the volume, model, horsepower, engine parameter and the like, the basic parameters of various vehicles are different, which affects the driving parameters of the vehicle, the driving speed and flexibility are different, the vehicle driving parameter is used to describe and measure the performance and characteristics of the vehicle in the driving process. Including the information of 100km acceleration, maximum speed, maximum horsepower, maximum torque and the like. By determining the vehicle driving parameter, support is provided for the next step of vehicle driving state prediction. For example, for a large volume vehicle, in the state that the driving speed exceeds a certain threshold, the driving direction should be straight, if the turn is completed, the speed needs to be reduced, to ensure the turn operation, according to the basic parameters and driving speed of the vehicle, the predicted driving direction of the vehicle can be determined.

[0047] According to the driving position information, the driving speed information and the vehicle driving parameter, the driving state prediction of the monitoring target vehicle is performed, and the vehicle predicted monitoring track is determined.

[0048] In one embodiment, it further includes:

[0049] According to the driving position information, the position area road extraction is performed through the electronic map system, the driving road information is obtained, including the road condition information and the passing road node;

[0050] According to the driving speed information and the vehicle driving parameter, the driving strategy is performed, the variation probability prediction is performed, and the driving direction prediction information is obtained.

[0051] According to the driving road information and the driving direction prediction information, the driving track prediction is performed.

[0052] According to the driving position information, driving speed information, and vehicle driving parameters, the driving state of the target vehicle is predicted. First, the position area road is extracted by the electronic map system based on the driving position information. The electronic map system is an application program based on computer technology and geographic information system (GIS) for displaying and managing geographic spatial data and providing navigation and positioning services. These systems usually contain a large amount of map data and geographic information, which can be used for navigation, travel planning, location search, geographic analysis, etc. For example: Gaode map, Baidu map, Tencent map, etc. The core of the electronic map system is map data, which can include road, building, terrain, river, city boundary, natural resources, etc. These data are usually obtained through geographic survey, remote sensing technology and other data collection methods, and stored in computer system. The driving road information is obtained, which includes road condition information and passing road nodes. The road condition information refers to the real-time traffic condition and road surface condition on the road, such as traffic congestion, road potholes, road wetness, road water, surrounding vehicle state, etc. The passing road node refers to the specific position or intersection point connecting different roads, road sections or traffic lines in the traffic network, which is usually a place with heavy traffic flow, such as crossroads, intersections, roundabouts, highway interchanges, etc.

[0053] According to the driving speed information and vehicle driving parameters, the driving strategy change probability prediction is performed. The driving strategy change probability prediction refers to the analysis of the driver's behavior and vehicle driving parameters to predict the probability of the driver changing the driving strategy in the future. For example, when the vehicle is in deceleration state and the front wheel direction of the vehicle deviates to the left, it indicates that the vehicle may change lane to the left or turn left; when the vehicle is in acceleration state and the front wheel direction of the vehicle deviates to the left, it indicates that the vehicle is going to overtake. The driving direction prediction information is obtained, which refers to the future driving direction of the vehicle.

[0054] Finally, the driving road information and the driving direction prediction information are fused, and the driving trajectory prediction is performed according to the fusion result. For example, when the front road surface has a traffic accident and the vehicles are all changing lane to the left, the vehicle is likely to change lane to the left; when the front road surface is a traffic intersection and the front vehicles are all turning left, the vehicle is likely to turn left at the traffic intersection. The vehicle prediction monitoring trajectory is obtained. By combining the driving road information and the driving direction prediction information, the accuracy of the vehicle prediction monitoring trajectory can be improved, thereby improving the accuracy of the vehicle intelligent tracking.

[0055] Continuously collect monitoring information of the monitoring target vehicle, and perform trajectory deviation calculation based on the vehicle predicted monitoring trajectory, perform tracking trajectory correction based on the trajectory deviation, and obtain vehicle tracking information.

[0056] In one embodiment, further comprising:

[0057] Based on the vehicle predicted monitoring trajectory and the monitoring information, determine the aiming point and the aiming distance;

[0058] Through the trajectory tracking algorithm module, perform trajectory tracking based on the aiming point and the aiming distance, and obtain vehicle trajectory calculation results;

[0059] Perform vehicle state monitoring through the traffic monitoring system, continuously collect and obtain monitoring information of the monitoring target vehicle, wherein the monitoring information includes vehicle position, vehicle speed, vehicle type, vehicle driving state, etc. Then, based on the monitoring information and the vehicle predicted monitoring trajectory, perform trajectory deviation calculation, which refers to quantitative analysis of the difference between the vehicle trajectory in a period of time and the reference trajectory.

[0060] First, based on the vehicle predicted monitoring trajectory and the monitoring information, determine the aiming point and the aiming distance, the aiming point refers to any point in the vehicle predicted monitoring trajectory, which can be set by those skilled in the art according to actual conditions, and the aiming distance refers to the distance from the vehicle to the aiming point obtained according to the current vehicle driving prediction trajectory. Then input the aiming point and the aiming position into the trajectory tracking algorithm module, which is embedded with a pure tracking algorithm, the pure tracking algorithm refers to a trajectory tracking control method based on the kinematic relationship between the vehicle position and the reference trajectory to realize vehicle lateral control, which has fast operation speed, strong real-time performance and good adaptability. Its principle is to establish an algorithm between the fixed parameter preview distance, the preview point and the vehicle position based on the geometric relationship between the vehicle and the reference path, calculate the steering radius and the driving curvature, and finally calculate the front wheel steering angle control quantity according to the known vehicle wheelbase and the preview distance. The vehicle can continuously approach the desired path under the action of the control quantity, and realize trajectory tracking. The advantage of pure tracking algorithm lies in its simplicity and real-time performance. Since it does not involve the complex process of target detection and recognition, it has fast speed and is suitable for real-time application. Through the use of pure tracking algorithm for trajectory tracking, vehicle trajectory calculation results are obtained.

[0061] In one embodiment, further comprising:

[0062] Extract the first target image of the monitoring information of the target vehicle;

[0063] Based on the aiming point, read the preview pixel number, and according to the aiming distance and the target vehicle, determine the center of the longest diameter pixel number as the center of the circle, and draw a target circle with the preview pixel number as the radius.

[0064] the intersection of the target circle and the vehicle predicted monitoring trajectory as a pre-aim pixel point;

[0065] a first actual motion direction of the pixel region is obtained, and an included angle between the first actual motion direction and the pre-aim pixel point is taken as a deviation angle;

[0066] a first actual turning angle is calculated based on the longest diameter pixel number, the pre-aim pixel number and the deviation angle;

[0067] a second target image of the target vehicle monitoring information is extracted, and a second actual turning angle is obtained by analyzing the second target image, wherein the first target image and the second target image are adjacent images;

[0068] the vehicle trajectory calculation result is obtained based on the first actual turning angle and the second actual turning angle.

[0069] In the trajectory tracking by using the pure tracking algorithm, first, a first target image of the target vehicle monitoring information is extracted, and the first target image is any one of the target vehicle monitoring information. Then, based on the aiming point, a pre-aim pixel number is read, the pre-aim pixel number is used to represent a distance, and a person skilled in the art can set it based on the actual situation, and the center of the longest diameter pixel number is taken as the center of the circle according to the aiming distance and the target vehicle, and a target circle is drawn with the pre-aim pixel number as the radius. And the intersection of the target circle and the vehicle predicted monitoring trajectory is taken as the pre-aim pixel point.

[0070] a first actual motion direction of the pixel region is obtained, and an included angle between the first actual motion direction and the pre-aim pixel point is taken as a deviation angle. Then a first actual turning angle is calculated based on the longest diameter pixel number, the pre-aim pixel number and the deviation angle. A second target image of the target vehicle monitoring information is extracted, and a second actual turning angle is obtained by analyzing the second target image, wherein the first target image and the second target image are adjacent images, and the adjacent images are images of adjacent image frames. Finally, the vehicle trajectory fitting is performed according to the first actual turning angle and the second actual turning angle, and the vehicle trajectory calculation result is obtained. By obtaining the vehicle trajectory calculation result, support is provided for the next step of calculating the vehicle trajectory deviation degree.

[0071] the trajectory deviation degree is obtained according to the vehicle trajectory calculation result and the vehicle predicted monitoring trajectory.

[0072] Specifically, the vehicle prediction monitoring trajectory is subtracted from the vehicle trajectory operation result to obtain a vehicle trajectory difference value, and the vehicle trajectory difference value is taken as a trajectory deviation degree. Finally, tracking trajectory correction is performed according to the trajectory deviation degree to obtain vehicle tracking information. Through the above method, the technical problem of low vehicle tracking accuracy according to the vehicle trajectory in the prior art is solved, and the accuracy and efficiency of vehicle tracking can be improved.

[0073] In one embodiment, as shown in Figure 3 An intelligent tracking system for vehicle trajectory is provided, comprising: a monitoring target vehicle determination module a, an information collection module b, a vehicle driving parameter determination module c, a vehicle prediction monitoring trajectory determination module d, a vehicle tracking information obtaining module e, wherein:

[0074] The monitoring target vehicle determination module a is used for vehicle state monitoring through a traffic monitoring system, vehicle feature traversal of a monitoring picture according to vehicle tracking features, and determination of a monitoring target vehicle.

[0075] The information collection module b is used for collection of vehicle information, driving position information, and driving speed information of the monitoring target vehicle.

[0076] The vehicle driving parameter determination module c is used for vehicle driving performance evaluation according to the vehicle information, and determination of vehicle driving parameters.

[0077] The vehicle prediction monitoring trajectory determination module d is used for driving state prediction of the monitoring target vehicle according to the driving position information, driving speed information, and the vehicle driving parameters, and determination of a vehicle prediction monitoring trajectory.

[0078] The vehicle tracking information obtaining module e is used for continuous collection of monitoring information of the monitoring target vehicle, trajectory deviation degree calculation based on the vehicle prediction monitoring trajectory, tracking trajectory correction based on the trajectory deviation degree, and obtaining of vehicle tracking information.

[0079] In one embodiment, the system further comprises:

[0080] A sample collection module is used for setting multiple types of tracking features, and performing sample collection of each type based on the multiple types of tracking features to obtain a multiple-type sample set.

[0081] A feature extraction module is used for feature extraction of the multiple-type sample set to obtain a feature set of each sample.

[0082] a sample feature traversal sequence setting module, configured to perform feature weight calculation on a feature set of each sample respectively, and set a sample feature traversal sequence based on the feature weight;

[0083] a traversal comparison module, configured to perform traversal comparison on the sample feature traversal sequence as vehicle tracking features.

[0084] In an embodiment, the system further comprises:

[0085] a feature salience determination module, configured to perform recognition degree analysis on features, and determine feature salience;

[0086] a feature repetition rate determination module, configured to perform repetition rate calculation on features, and determine feature repetition rate;

[0087] a rule-constrained feature determination module, configured to perform tracking rule analysis based on multiple types of tracking features, and determine rule-constrained features;

[0088] a feature evaluation value obtaining module, configured to perform feature evaluation on the feature salience, feature repetition rate, and rule-constrained features according to a preset weight setting value, and obtain a feature evaluation value;

[0089] a weight obtaining module, configured to take the feature evaluation value as a weight coefficient, and obtain the weight of each feature.

[0090] In an embodiment, the system further comprises:

[0091] an information obtaining module, configured to perform location area road extraction through an electronic map system according to the driving location information, and obtain driving road information, including road condition information and passing road nodes;

[0092] a driving direction prediction information obtaining module, configured to perform driving strategy and variation probability prediction according to the driving speed information and vehicle driving parameters, and obtain driving direction prediction information;

[0093] a driving trajectory prediction module, configured to perform driving trajectory prediction according to the driving road information and the driving direction prediction information.

[0094] In an embodiment, the system further comprises:

[0095] a targeting point and targeting distance determination module, configured to determine a targeting point and a targeting distance based on the vehicle prediction monitoring trajectory and monitoring information;

[0096] a vehicle trajectory operation result obtaining module configured to obtain a vehicle trajectory operation result by a trajectory tracking algorithm module based on the aiming point and the aiming distance;

[0097] a trajectory deviation degree obtaining module configured to obtain a trajectory deviation degree according to the vehicle trajectory operation result and the vehicle predicted monitoring trajectory.

[0098] In one embodiment, the system further comprises:

[0099] a first target image extraction module configured to extract a first target image of monitoring information of the target vehicle;

[0100] a target circle drawing module configured to read a pre-aiming pixel number based on the aiming point, and determine a center of a longest diameter pixel number as a circle center according to the aiming distance and the target vehicle, and draw a target circle with the pre-aiming pixel number as a radius;

[0101] a pre-aiming pixel point obtaining module configured to take an intersection of the target circle and the vehicle predicted monitoring trajectory as a pre-aiming pixel point;

[0102] a deviation angle obtaining module configured to obtain a first actual motion direction of the pixel region, and take an included angle between the first actual motion direction and the pre-aiming pixel point as a deviation angle;

[0103] a first actual turning angle calculation module configured to calculate a first actual turning angle based on the longest diameter pixel number, the pre-aiming pixel number and the deviation angle;

[0104] a second actual turning angle obtaining module configured to extract a second target image of the monitoring information of the target vehicle, and analyze the second target image to obtain a second actual turning angle, wherein the first target image and the second target image are adjacent images;

[0105] a vehicle trajectory operation result obtaining module configured to obtain the vehicle trajectory operation result based on the first actual turning angle and the second actual turning angle.

[0106] In summary, the intelligent tracking method and system for vehicle trajectory provided by the present application have the following technical effects:

[0107] 1. The technical problem of low accuracy of vehicle tracking according to vehicle trajectory in the prior art is solved, and the accuracy and efficiency of vehicle tracking can be improved.

[0108] 2. By combining the driving road information and the driving direction prediction information to predict the driving track, the accuracy of the vehicle prediction monitoring track can be improved, thereby improving the accuracy of the vehicle intelligent tracking.

[0109] 3. The pure tracking algorithm has the advantages of simplicity and real-time. Since it does not involve the complex process of target detection and recognition, it is faster and suitable for real-time applications

[0110] The technical features of the above embodiments can be combined in any manner. In order to make the description concise, all possible combinations of the technical features in the above embodiments are not described, but as long as the combinations of the technical features do not contradict, they should be considered within the scope of the present disclosure.

[0111] The above embodiments only express several implementation manners of the present application, and the description is specific and detailed, but it should not be understood as a limitation on the scope of the patent. It should be pointed out that for ordinary skilled persons in the art, some modifications and improvements can be made without departing from the concept of the present application, and these are within the protection scope of the present application. Therefore, the protection scope of the patent of the present application should be subject to the appended claims.

Claims

1. An intelligent tracking method of a vehicle trajectory, characterized in that, The application relates to a vehicle state monitoring method based on a traffic monitoring system. The vehicle state monitoring method comprises the following steps: Vehicle state monitoring is performed through a traffic monitoring system, vehicle feature traversal of a monitoring picture is performed according to vehicle tracking features, and a monitoring target vehicle is determined; Vehicle information, driving position information and driving speed information of the monitoring target vehicle are collected; Vehicle driving performance evaluation is performed according to the vehicle information, and vehicle driving parameters are determined; Driving state prediction of the monitoring target vehicle is performed according to the driving position information, the driving speed information and the vehicle driving parameters, and a vehicle predicted monitoring track is determined; Monitoring information of the monitoring target vehicle is continuously collected, track deviation degree calculation is performed based on the vehicle predicted monitoring track, tracking track correction is performed based on the track deviation degree, and vehicle tracking information is obtained; The vehicle feature traversal of the monitoring picture is performed according to the vehicle tracking features, and the monitoring target vehicle is determined, which comprises the following steps: Multiple types of tracking features are set, and multiple type sample sets are obtained through sample collection based on the multiple types of tracking features; Feature extraction is performed on the multiple type sample sets, and feature sets of the samples are obtained; Feature weight calculation is performed on the feature sets of the samples respectively, a sample feature traversal sequence is set based on the feature weights; The sample feature traversal sequence is used as the vehicle tracking features for traversal comparison; The track deviation degree calculation comprises the following steps: Aiming point and aiming distance are determined based on the vehicle predicted monitoring track and the monitoring information; Track tracking is performed based on the aiming point and the aiming distance through a track tracking algorithm module, and vehicle track operation results are obtained; Track deviation degree is obtained according to the vehicle track operation results and the vehicle predicted monitoring track; The vehicle track operation results are obtained through the following steps: A first target image of the monitoring information of the target vehicle is extracted; Aiming point is used to read a pre-aiming pixel number, a center of a longest diameter pixel number is determined as a circle center according to the aiming distance and the target vehicle, a target circle is drawn with the pre-aiming pixel number as a radius; An intersection of the target circle and the vehicle predicted monitoring track is used as a pre-aiming pixel point; A first actual motion direction of a pixel region is obtained, and an included angle between the first actual motion direction and the pre-aiming pixel point is used as a deviation angle; A first actual turning angle is calculated based on the longest diameter pixel number, the pre-aiming pixel number and the deviation angle; A second target image of the monitoring information of the target vehicle is extracted, and a second actual turning angle is obtained by analyzing the second target image, wherein the first target image and the second target image are adjacent images; 2. The method of claim 1, wherein, The vehicle track operation results are obtained based on the first actual turning angle and the second actual turning angle. The feature weight calculation on the feature sets of the samples comprises the following steps: Feature obviousness is determined through feature recognition degree analysis; Feature repetition rate is calculated, and feature repetition rate is determined; Rule constraint features are determined through tracking rule analysis based on the multiple types of tracking features; Feature evaluation is performed on the feature obviousness, the feature repetition rate and the rule constraint features according to preset weight setting values, and feature evaluation values are obtained; The feature evaluation values are used as weight coefficients, and weight values of the features are obtained.

3. The method of claim 1, wherein, According to the driving position information, the driving speed information, and the vehicle driving parameter, the driving state of the target vehicle is predicted, and a vehicle prediction monitoring track is determined, including: According to the driving position information, a position area road is extracted through an electronic map system to obtain driving road information, including road condition information and passing road nodes; According to the driving speed information and the vehicle driving parameter, a driving strategy is determined, and a variation probability is predicted to obtain driving direction prediction information; According to the driving road information and the driving direction prediction information, a driving track is predicted.

4. An intelligent tracking system for vehicle trajectory, characterized in that, Steps for performing any one of the intelligent tracking methods of the vehicle track in claims 1-3, including: A target vehicle determination module is configured to monitor the vehicle state through a traffic monitoring system, traverse the vehicle features of the monitoring picture according to the vehicle tracking features, and determine the target vehicle; An information collection module is configured to collect the vehicle information, driving position information, and driving speed information of the target vehicle; A vehicle driving parameter determination module is configured to evaluate the vehicle driving performance according to the vehicle information, and determine the vehicle driving parameter; A vehicle prediction monitoring track determination module is configured to predict the driving state of the target vehicle according to the driving position information, the driving speed information, and the vehicle driving parameter, and determine the vehicle prediction monitoring track; A vehicle tracking information acquisition module is configured to continuously collect the monitoring information of the target vehicle, calculate the track deviation degree based on the vehicle prediction monitoring track, correct the tracking track based on the track deviation degree, and obtain the vehicle tracking information.

Images