Urban traffic monitoring and early warning method, device and medium based on artificial intelligence

Abstract

The present application provides an artificial intelligence-based urban traffic monitoring and early warning method, device and medium, which are applicable to an urban traffic monitoring center platform. The method comprises: performing monitoring and early warning configuration processing on an urban traffic monitoring task; in response to the execution of the urban traffic monitoring task, collecting driving video data of vehicles or people passing through a traffic node area through an urban traffic monitoring camera, performing trajectory feature processing on each traffic driving image of the vehicle or person, and obtaining the traffic driving trajectory features of each traffic driving image of the vehicle or person; determining an early warning verification image of an area corresponding to the traffic node according to the traffic driving trajectory features and image acquisition quality information of the vehicle or person in each traffic driving image; if the early warning verification image of the vehicle or person in the area corresponding to the traffic node meets the monitoring and early warning rules, sending early warning notification information to an early warning terminal; the scheme can improve the real-time performance of urban traffic monitoring and early warning.

Description

Technical Field

[0001] The present application relates to the fields of information technology and artificial intelligence technology, and more specifically, to an urban traffic monitoring and early warning method, device and medium based on artificial intelligence. Background Art

[0002] Traditional urban traffic monitoring systems mainly rely on manual monitoring and simple automation technology, and there is a problem of insufficient proactive prevention. Most urban traffic cameras only have the function of collecting evidence after an accident occurs, and cannot play a role in preventing and warning suspicious and abnormal urban traffic behaviors, making it difficult to meet the needs of urban traffic safety. With the development of technologies such as artificial intelligence and big data and the strong demand for urban traffic safety from all walks of life, traditional simple and passive urban traffic monitoring can no longer meet the needs of daily diversified urban traffic scenarios and safety.

[0003] Therefore, how to improve the real-time performance of urban traffic monitoring is a technical problem that technicians in this field need to solve. Summary of the invention

[0004] The purpose of this application is to provide an urban traffic monitoring and early warning method, device and a computer-readable storage medium based on artificial intelligence, which improves the real-time performance of urban traffic monitoring.

[0005] To achieve the above objectives, the first aspect of the present application provides an urban traffic monitoring and early warning method based on artificial intelligence, which is applicable to an urban traffic monitoring center platform. The urban traffic monitoring and early warning method includes:

[0006] Perform monitoring and early warning configuration processing on the urban traffic monitoring task to obtain the monitoring and early warning rules of the urban traffic monitoring task;

[0007] Obtaining a traffic camera identification set for urban traffic control, wherein the traffic camera identification set indicates a plurality of urban traffic monitoring cameras, and the urban traffic monitoring cameras are deployed at key traffic nodes in the city and are used to collect images of vehicles and pedestrians;

[0008] In response to the execution of the urban traffic monitoring task, the urban traffic monitoring camera collects driving video data of vehicles or people passing through the traffic node area, wherein the driving video data includes at least two traffic driving images of the vehicles or people;

[0009] Performing trajectory feature processing on each traffic driving image of the vehicle or person to obtain traffic driving trajectory features of each traffic driving image of the vehicle or person;

[0010] Determine a warning verification image of the area corresponding to the traffic node according to the traffic trajectory characteristics and image acquisition quality information of the vehicle or person in each traffic driving image, wherein the image acquisition quality information is used to measure the quality of the vehicle or person in the traffic driving image acquired by the traffic monitoring camera;

[0011] If the warning verification image of the vehicle or person in the area corresponding to the traffic node meets the monitoring warning rule, a warning notification message is sent to the warning terminal.

[0012] To achieve the above-mentioned purpose, the second aspect of the present application provides an urban traffic monitoring and early warning device based on artificial intelligence, which is used in an urban traffic monitoring center platform. The urban traffic monitoring and early warning device includes:

[0013] The configuration module is used to perform monitoring and early warning configuration processing on the urban traffic monitoring task and obtain the monitoring and early warning rules of the urban traffic monitoring task;

[0014] An acquisition module is used to acquire a traffic camera identification set for urban traffic control, wherein the traffic camera identification set indicates a plurality of urban traffic monitoring cameras, and the urban traffic monitoring cameras are deployed at key traffic nodes in the city and are used to collect images of vehicles and pedestrians;

[0015] A collection module, configured to collect driving video data of vehicles or people passing through the traffic node area through the city traffic monitoring camera in response to the execution of the city traffic monitoring task, wherein the driving video data includes at least two traffic driving images of the vehicle or the person;

[0016] A feature processing module, used for performing trajectory feature processing on each traffic driving image of the vehicle or person to obtain a traffic driving trajectory feature of each traffic driving image of the vehicle or person;

[0017] An early warning processing module is used to determine the early warning verification image of the area corresponding to the traffic node based on the traffic driving trajectory characteristics and the image acquisition quality information of the vehicle or person in each traffic driving image, and the image acquisition quality information is used to measure the quality of the vehicle or person in the traffic driving image collected by the traffic monitoring camera; if the early warning verification image of the vehicle or person in the area corresponding to the traffic node meets the monitoring early warning rules, a warning notification information is sent to the early warning terminal.

[0018] To achieve the above-mentioned purpose, the fourth aspect of the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned urban traffic monitoring and early warning method are implemented.

[0019] It can be seen from the above scheme that the urban traffic monitoring and early warning method based on artificial intelligence provided by the present application is applicable to the urban traffic monitoring center platform, and the urban traffic monitoring and early warning method includes: performing monitoring and early warning configuration processing on the urban traffic monitoring task to obtain the monitoring and early warning rules of the urban traffic monitoring task; obtaining the traffic camera identification set of the urban traffic control, the traffic camera standard head identification set indicates multiple urban traffic monitoring cameras, and the urban traffic monitoring cameras are deployed at key traffic nodes in the city for image acquisition of vehicles and pedestrians; in response to the execution of the urban traffic monitoring task, the driving video data of vehicles or people passing through the traffic node area is collected by the urban traffic monitoring camera, and the driving video data includes at least two traffic driving images of the vehicle or the person; The method comprises the following steps: performing trajectory feature processing on each traffic driving image of a vehicle or a person to obtain the traffic driving trajectory feature of each traffic driving image of the vehicle or the person; determining the warning verification image of the corresponding area of ​​the traffic node according to the traffic driving trajectory feature and the image acquisition quality information of the vehicle or the person in each traffic driving image, and the image acquisition quality information is used to measure the quality of the vehicle or the person in the traffic driving image acquired by the traffic monitoring camera; if the warning verification image of the vehicle or the person in the corresponding area of ​​the traffic node meets the monitoring warning rule, sending warning notification information to the warning terminal; the scheme can screen out high-quality warning verification images from the traffic driving images of the vehicle or the person acquired by the camera, and perform urban traffic warning based on the warning verification images, which can improve the real-time performance of urban traffic monitoring and warning. In addition, since the scheme can select high-quality warning verification images based on traffic driving trajectory features and image acquisition quality information, the accuracy of urban traffic monitoring and warning can be improved.

[0020] It should be understood that the foregoing general description and the following detailed description are exemplary only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following is a brief introduction to the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. The drawings are used to provide a further understanding of the present disclosure and constitute a part of the specification. Together with the following specific implementation methods, they are used to explain the present disclosure, but do not constitute a limitation to the present disclosure. In the drawings:

[0022] Figure 1 An architectural diagram of an urban traffic monitoring and early warning system based on artificial intelligence provided in an embodiment of the present application;

[0023] Figure 2a A flowchart of an artificial intelligence-based urban traffic monitoring and early warning method provided in an embodiment of the present application;

[0024] Figure 2b A flowchart of an image quality verification method provided in an embodiment of the present application;

[0025] Figure 3 A flow chart of a method for updating a traffic monitoring image database provided in an embodiment of the present application;

[0026] Figure 4 A structural diagram of an urban traffic monitoring and early warning device based on artificial intelligence provided in an embodiment of the present application;

[0027] Figure 5 A structural diagram of a server provided in an embodiment of the present application. DETAILED DESCRIPTION

[0028] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0029] Artificial Intelligence (AI) is a branch of computer science that aims to create machines or software systems that can perform human intelligent activities. These activities include but are not limited to perception, reasoning, learning, planning, communication, and problem solving. Artificial intelligence systems are able to simulate human cognitive functions such as visual recognition, language understanding, decision making, and predictive analysis to perform tasks or provide decision support in various complex environments.

[0030] The development of AI technology involves multiple fields, including machine learning, natural language processing, computer vision, and robotics. Machine learning is a core component of AI that enables computer systems to learn and improve from data without being explicitly programmed. Natural language processing enables machines to understand and generate human language, while computer vision enables machines to recognize and process visual information from images or multi-dimensional data. As technology advances, AI systems are becoming more sophisticated, and their applications in multiple industries such as healthcare, finance, transportation, and education are becoming more widespread, greatly improving efficiency, reducing costs, and providing new solutions to some of the most pressing global problems.

[0031] Traditional urban traffic monitoring systems mainly rely on manual monitoring and simple automation technology, which cannot actively realize urban traffic monitoring and early warning, and cannot meet the real-time nature of urban traffic monitoring and early warning. In addition, these systems have the following limitations: First, traditional monitoring systems are highly dependent on the quality of image acquisition, but in complex environments, such as at night or in bad weather conditions, the image quality is often poor, resulting in a significant reduction in monitoring effects. In addition, these systems usually lack effective image preprocessing and target feature extraction technology, and cannot accurately identify and track target objects, such as vehicles and pedestrians, resulting in low warning accuracy, and thus unable to provide effective decision support for traffic management.

[0032] Therefore, this application uses artificial intelligence and other technologies to provide a clustered urban traffic area integrated monitoring and early warning system, the main functions of which include active real-time capture and display of vehicles or people in urban traffic monitoring videos, timely deployment and early warning of controlled vehicles or people, dynamic trajectory tracking, etc. In this application, the deployed traditional urban traffic cameras can be reused, that is, the input video stream of the urban traffic monitoring and early warning system can be the video collected by traditional urban traffic cameras, and no additional deployment operation or maintenance costs are required. Since the monitoring system adopts a clustered design mode, horizontal expansion can be achieved by simply adding new traffic monitoring nodes, which is conducive to the expansion of the monitoring data volume, and can support the traffic monitoring and early warning needs of traffic monitoring areas, traffic monitoring administrative districts, and cities. It can be seen that this application can provide an integrated urban traffic monitoring and early warning solution for relevant departments and institutions.

[0033] In order to facilitate understanding of the urban traffic monitoring and early warning method based on artificial intelligence provided by this application, the system used is introduced below. Figure 1 , which shows an architecture diagram of a city traffic monitoring and early warning system provided by an embodiment of the present application, such as Figure 1 As shown, it includes a traffic camera 100, a city traffic monitoring center platform 200 and an early warning terminal 300. The traffic camera 100 and the city traffic monitoring center platform 200, and the city traffic monitoring center platform 200 and the early warning terminal 300 are connected through a network.

[0034] In specific implementation, the urban traffic management agency can deploy a list of traffic cameras that need to be controlled at the key traffic nodes in the city, that is, deploy traffic cameras 100, which are used to collect image streams and video streams of vehicles or people in the corresponding areas of key traffic nodes. They can include high-definition cameras, smart cameras, etc. in various application scenarios, which are not specifically limited here, such as customized pedestrian or vehicle capture machines. For video streams, it is necessary to use video decoding services for decoding so that the urban traffic monitoring center platform 200 can identify the target vehicles or pedestrians therein. When expanding traffic monitoring points, it is only necessary to connect the traffic cameras at the newly added monitoring points to the system.

[0035] The city traffic monitoring center platform 200 is used to process the images or videos input by the traffic camera 100 to realize active monitoring and early warning. The city traffic monitoring center platform 200 can provide services such as control and early warning for traffic management agencies. It is understandable that in order to improve the business processing capability, the city traffic monitoring center platform 200 can be implemented by a server, or a cluster design can be adopted, using LB (Load Balance) technology to share network services, network traffic and other tasks to multiple network node devices or multiple links of the cluster, thereby ensuring the high reliability of the business. In this embodiment, each server node can be responsible for the monitoring task of a certain area, that is, it is responsible for the images collected by the traffic camera 100 in a certain camera list. When expanding the entire city traffic monitoring system, it is only necessary to set the corresponding server node for the newly added traffic monitoring area, which is conducive to the expansion of the monitoring data magnitude.

[0036] The early warning terminal 300 can be a mobile terminal such as a mobile phone or a fixed terminal such as a PC (Chinese full name: personal computer, English full name: personal computer), which can receive early warning notification information sent by the urban traffic monitoring center platform 200 in real time. In addition, monitoring early warning rules can be deployed on the urban traffic monitoring center platform 200, and database maintenance can be performed.

[0037] The embodiment of the present application discloses a city traffic monitoring and early warning method, which improves the real-time performance of city traffic monitoring.

[0038] See also Figure 2a , a flow chart of a city traffic monitoring and early warning method provided in an embodiment of the present application, such as Figure 2a As shown, including:

[0039] S101: The management terminal sends demand information for urban traffic monitoring tasks to the urban traffic monitoring center platform, and the demand information may include a set of traffic camera identifications and warning description information for traffic monitoring; wherein the traffic camera identification set includes camera identifications corresponding to multiple urban traffic monitoring cameras, and the urban traffic monitoring cameras are deployed at key traffic nodes in the city for capturing images of vehicles and pedestrians.

[0040] In the specific implementation, the manager can configure the demand information of this urban traffic monitoring task through the interactive interface of the management terminal, such as configuring the traffic nodes or areas to be monitored, the corresponding urban traffic monitoring cameras, the trigger conditions and methods of early warning, etc.

[0041] For example, if the monitored object is a vehicle, the demand information of the urban traffic monitoring task here can be determined based on the monitored object, that is, the vehicle in the captured image meets the pre-set warning conditions. The warning conditions here may include the capture location, capture time, vehicle information, etc.

[0042] S102: The urban traffic monitoring center platform performs monitoring and early warning configuration processing on the urban traffic monitoring task to obtain monitoring and early warning rules for the urban traffic monitoring task.

[0043] After receiving the demand information, the urban traffic monitoring center platform can configure the monitoring warning rules adapted to the configuration, wherein the monitoring warning rules include: monitoring objects such as vehicles or people, warning conditions, i.e., under what conditions the warning is triggered, warning methods, i.e., what methods are used for warning, etc. In one embodiment, the warning conditions may include whether the image captured by the urban traffic camera matches the image in the database, etc. The warning conditions or monitoring warning rules can be configured according to actual needs.

[0044] S103: The city traffic monitoring center platform triggers the city traffic monitoring task, and collects driving video data of vehicles or people passing through the traffic node area through the city traffic monitoring camera, and the driving video data includes at least two traffic driving images of the vehicle or the person.

[0045] Urban traffic monitoring cameras collect driving video data of vehicles or people; wherein the driving video data includes multiple traffic driving images containing the vehicles or people.

[0046] In this step, the urban traffic monitoring camera collects traffic images containing vehicles or people, and multiple traffic driving images constitute the driving video data of the vehicle or person. Taking a vehicle as an example, when a vehicle is driving in the monitoring area corresponding to a traffic node, as the vehicle moves, the urban traffic monitoring camera can track and capture the vehicle to obtain multiple traffic driving images of the vehicle to form the driving video data of the vehicle.

[0047] For vehicles or pedestrians, a face detection algorithm or a target recognition algorithm can also be used to determine a face frame containing a face, or a vehicle marker frame of a vehicle. For example, the input of a face detection algorithm is a picture, and the output is a sequence of face frame coordinates. Generally, the output face coordinate frame is a square facing upwards, but it can also be a rectangle facing upwards or a rectangle with a rotation direction.

[0048] For vehicles, this embodiment does not limit the specific recognition algorithm. Multiple features of pixel points in vehicle traffic images can be extracted through a convolutional neural network or a deep learning network. All pixel points in the target image are classified based on the extracted features and divided into vehicle areas and background areas, and finally the area corresponding to the target object is obtained.

[0049] S104: The urban traffic monitoring camera collects driving video data of vehicles or people passing through the traffic node area and sends it to the urban traffic monitoring center platform;

[0050] S105: The urban traffic monitoring center platform performs trajectory feature processing on each traffic driving image of the vehicle or person to obtain traffic driving trajectory features of each traffic driving image of the vehicle or person.

[0051] Among them, the traffic trajectory characteristics of vehicles or people refer to a series of descriptive parameters extracted by analyzing the movement paths of vehicles or pedestrians in a certain time and space in the urban traffic monitoring system. These characteristics can reflect the driving behavior patterns of individuals, including but not limited to speed changes, driving direction, acceleration, braking frequency, degree of tortuosity of driving routes, and time spent in specific areas. Through the data collected by traffic cameras and other sensors, these trajectory characteristics can be quantitatively analyzed to identify the difference between normal driving patterns and abnormal behaviors.

[0052] In this step, deep learning models, especially convolutional neural networks (CNNs), can be used to implement traffic trajectory feature processing. Deep learning models can be trained to recognize specific trajectory patterns and compare these patterns with known traffic rules or normal behavior benchmarks to identify abnormal behaviors. In addition, using machine learning algorithms such as hidden Markov models (HMMs) or long short-term memory networks (LSTMs), time series analysis of trajectory data can be performed to predict future movement trends and behaviors.

[0053] In this step, key information can also be extracted from the continuous frames of traffic driving video data. Using computer vision techniques such as background subtraction, optical flow estimation or deep learning algorithms, the system can identify and track the position changes of vehicles or pedestrians at different time points. Subsequently, by analyzing these position changes, dynamic characteristics such as speed, acceleration, and driving direction can be calculated.

[0054] S106: Determine a warning verification image of the area corresponding to the traffic node according to the traffic driving trajectory characteristics and the image acquisition quality information of the vehicle or person in each traffic driving image; the image acquisition quality information is used to measure the quality of the vehicle or person in the traffic driving image captured by the traffic monitoring camera.

[0055] In this step, the urban traffic monitoring center platform conducts a quality assessment on the vehicles or people in each traffic driving image, and determines the early warning verification image of the corresponding area of ​​the traffic node based on the trajectory characteristics of the traffic driving image and the image acquisition quality information of the vehicles or people in each traffic driving image. In order to ensure the accuracy of monitoring and early warning, the quality requirements for the target objects, i.e., vehicles or people, in the early warning verification image are relatively high.

[0056] The image acquisition quality information is used to measure the quality of the vehicle or person in the traffic driving image captured by the traffic monitoring camera. The image acquisition quality information can be expressed in various forms, such as an image acquisition quality score. The higher the score, the higher the image acquisition quality, and vice versa. The image acquisition quality may include the quality of the vehicle or person in the captured traffic driving image, including but not limited to the image clarity parameters, image position information, image brightness information, image proportion information, and camera angle parameters when the traffic monitoring camera captures the vehicle or person in the traffic driving image.

[0057] It should be noted that the present embodiment does not limit the specific calculation method of the image acquisition quality score. For example, it can be calculated according to the quality condition parameter value of the vehicle or person and the preset image quality calculation function. The quality condition parameter value is the value of the condition parameter, and the quality condition parameter may include the blurriness of the pedestrian (or face) image, the three-dimensional deflection angle of the pedestrian or vehicle, the brightness of the pedestrian or vehicle image, or the area of ​​the pedestrian image. Among them, the three-dimensional deflection angle of the pedestrian or vehicle may refer to the pitch angle (pitch), yaw angle (yaw) and roll angle (roll) of the vehicle, etc., then the condition parameter value may include the blurriness value, the three-dimensional deflection angle value, the brightness value or the area value, etc. The image quality calculation function is determined according to the pedestrian or vehicle recognition accuracy distribution and / or the confidence distribution corresponding to the condition parameters of the pedestrian or vehicle.

[0058] In a specific implementation, the recognition accuracy distribution or confidence distribution corresponding to the quality condition parameter can be fitted with a data curve to obtain a fitting function, and then the quality calculation function is determined according to the fitting function. The recognition accuracy distribution corresponding to the condition parameter can refer to the distribution of the recognition accuracy with respect to blur, the distribution of the recognition accuracy with respect to the three-dimensional deflection angle, the distribution of the recognition accuracy with respect to brightness, or the distribution of the recognition accuracy with respect to the area of ​​the target area; the confidence distribution corresponding to the condition parameter can refer to the distribution of the confidence with respect to blur, the distribution of the confidence with respect to the three-dimensional deflection angle, the distribution of the confidence with respect to brightness, or the distribution of the confidence with respect to the area of ​​the target area.

[0059] The image acquisition quality scores of vehicles or people in all traffic driving images in the driving video data can be calculated, and the traffic driving image with the highest image acquisition quality score can be selected as the warning verification image in combination with the traffic driving trajectory characteristics.

[0060] In order to improve the accuracy of urban traffic warnings, the traffic driving trajectory of the vehicle or person in the area corresponding to the traffic node can be generated according to the traffic driving trajectory characteristics; and the warning verification image of the area corresponding to the traffic node can be determined according to the traffic driving trajectory and the image acquisition quality information.

[0061] Specifically, in order to improve efficiency and accuracy, the traffic driving image to be verified is extracted from the driving video data in video time sequence; if the traffic driving trajectory is within a first preset trajectory range and the image acquisition quality score is greater than or equal to a first preset threshold, the traffic driving image to be verified is determined as a warning verification image, wherein the first preset trajectory range represents that the vehicle or person is in a normal driving state in the traffic node area; if the traffic driving trajectory is within the first preset trajectory range and the image acquisition quality score is less than the first preset threshold, the step of extracting the traffic driving image to be verified from the driving video data in video time sequence is re-executed; if the traffic driving trajectory is within the second preset trajectory range, the traffic driving image with the highest image acquisition quality score is selected as the warning verification image, wherein the second preset trajectory range represents that the vehicle or person is in an abnormal driving state in the traffic node area.

[0062] In order to improve the efficiency of urban traffic monitoring and early warning, the image acquisition quality score of vehicles or people in each traffic driving image can be calculated accordingly. As long as there is a traffic driving image whose image acquisition quality score needs to be greater than the set first preset value and the traffic driving trajectory meets the requirements, it will be used as a warning verification image without calculating the subsequent traffic driving images. That is, this step may include: determining the current traffic driving image in the driving video data, calculating the image acquisition quality score of the vehicle or person in the current traffic driving image; if the image acquisition quality score is greater than or equal to the first preset value, and the driving plan meets the preset requirements, then determining the current traffic driving image as the warning verification image; if the image acquisition quality score is less than the first preset value, then re-entering the step of determining the current traffic driving image in the video data; if there is no traffic driving image with a quality score greater than or equal to the first preset value in the trajectory data, then determining the traffic driving image with the highest image acquisition quality score as the warning verification image. It can be understood that the first preset value is not limited here, and the user can make timely adjustments according to the actual monitoring situation, such as monitoring accuracy.

[0063] To improve the accuracy of monitoring and warning, if the traffic trajectory is within the second preset trajectory range, the traffic image with the highest image acquisition quality score and the traffic trajectory characteristics of the image are used to comprehensively determine whether the image can be used as a warning verification. Figure 2b , in one embodiment, may include:

[0064] S1061. If the traffic driving trajectory is within a second preset trajectory range, obtain a target traffic driving image with a highest image acquisition quality score.

[0065] If the traffic trajectory is within the second preset trajectory range, it indicates that the vehicle or person is in an abnormal driving state in the traffic node area. At this time, the quality requirements for image selection are very high to prevent false recognition of image quality (that is, the image with the highest score is selected, but the actual image quality is not high), then it is necessary to further confirm the quality score of the image. In order to be close to the actual scene, the embodiment of the present application can be combined with the driving trajectory to further verify the image with the highest score. The specific calculation of the comprehensive quality score of the image is further verified.

[0066] S1062. Combining the image acquisition quality score of the target traffic driving image and the multiple traffic driving trajectory features corresponding to the traffic driving trajectory into an image quality indicator sequence, wherein the image quality indicator sequence includes at least one image quality indicator, wherein one image quality indicator corresponds to one image acquisition quality score or one traffic driving trajectory feature.

[0067] The urban traffic monitoring center platform can regard each trajectory feature and image acquisition quality score as an indicator to measure image quality.

[0068] S1063. Calculate the standard deviation of each image quality indicator in the image quality indicator sequence to obtain the indicator standard deviation corresponding to each image quality indicator in the image quality indicator sequence, wherein the indicator standard deviation measures the degree of discreteness of the indicators in the image quality indicator sequence.

[0069] The urban traffic monitoring center platform can calculate the standard deviation of each indicator. The standard deviation can measure the degree of dispersion of the data. The larger the standard deviation, the greater the variability of the indicator, and the greater the impact on the comprehensive score of image quality.

[0070] S1064. Calculate the correlation coefficient of the image quality indicator sequence to obtain the indicator correlation coefficient between the image quality indicators, wherein the indicator correlation coefficient is used to measure the linear relationship between the image quality indicators.

[0071] The Urban Traffic Monitoring Center platform can calculate the correlation coefficients between various indicators to evaluate the linear relationship between them, thereby obtaining the correlation coefficient matrix corresponding to the sequence. The correlation coefficient matrix can help the machine identify which indicators are independent of each other and which are highly correlated.

[0072] S1065. Determine a weight corresponding to each image quality indicator in the image quality indicator sequence according to an indicator standard deviation corresponding to each image quality indicator and an indicator correlation coefficient between the image quality indicators.

[0073] The Urban Traffic Monitoring Center Platform can calculate the information content of each indicator based on its standard deviation and correlation coefficient. The greater the information content, the more unique the information provided by the indicator, and the greater its weight should be; the weight of each indicator is calculated based on its information content. The weight can be obtained by dividing the information content of each indicator by the sum of the information content of all indicators.

[0074] In one embodiment, the amount of information may be calculated using the following formula:

[0075]

[0076] Among them, C j is the information content of the jth indicator, σ j is the standard deviation of the jth indicator, r ij is the correlation coefficient between the ith indicator and the jth indicator, and n is the total number of indicators.

[0077] In one embodiment, the weight of the indicator can be calculated by the following formula:

[0078]

[0079] Among them, w j is the weight of the jth indicator, C j is the information content of the jth indicator, the denominator It is the sum of all indicator information.

[0080] S1066: Perform weighted processing on the image quality indicator sequence according to the weight corresponding to each image quality indicator in the image quality indicator sequence to obtain a comprehensive quality score of the traffic driving image to be verified.

[0081] The urban traffic monitoring center platform can multiply the score of each indicator by its corresponding weight, and then sum the weighted scores of all indicators to obtain a comprehensive score for the target traffic driving image.

[0082] S1067: Determine, according to the comprehensive quality score, that the target traffic driving image is a warning verification image.

[0083] If the target traffic image is greater than a preset threshold, the target traffic image is determined to be a warning verification image, indicating that the image with the highest image acquisition quality score has a high confidence level and can be used as a warning verification image. Otherwise, the remaining traffic images need to be selected as warning verification images.

[0084] It can be seen that when the vehicle or behavior traffic trajectory is in an abnormal state, the embodiment of the present application can comprehensively judge whether the image can be used as an early warning verification based on the traffic driving image with the highest image acquisition quality score and the traffic driving trajectory characteristics of the image, and perform secondary verification to improve the accuracy of the early warning.

[0085] S106: If the warning verification image of the vehicle or person in the area corresponding to the traffic node meets the monitoring warning rule, a warning notification message is sent to the warning terminal.

[0086] In this step, the urban traffic monitoring center platform determines whether the warning verification image meets the monitoring warning rules during the execution of the urban traffic monitoring task. If so, it sends a warning notification message to the warning terminal.

[0087] Exemplarily, this step may include: matching the warning verification image of the vehicle or person with the image in the traffic monitoring image database, and if the matching result meets the preset warning condition, sending warning notification information to the warning terminal.

[0088] In one embodiment, in order to improve the accuracy and efficiency of early warning, if the matching result indicates a successful match, the regional level corresponding to the traffic node is queried in the traffic monitoring database, and the traffic warning level of the vehicle or person is determined according to the regional level; and the early warning notification information corresponding to the traffic warning level is sent to the early warning terminal. That is, different levels of early warning are used for different traffic area levels.

[0089] In a specific implementation, if the traffic monitoring object is a portrait, a face image can be converted into a string of fixed-length values. This string of values ​​is called a face feature, which has the ability to characterize the characteristics of the face. The face feature extraction algorithm will align the face to a predetermined pattern based on the coordinates of the key points of the facial features, and then calculate the features. The face feature extraction algorithm is used to identify the identity corresponding to the input face image. Its input is a face feature, which is compared one by one with the features corresponding to N images in the data image library to find the feature with the highest similarity to the input feature. This highest similarity value is compared with the preset threshold. If it is greater than the threshold, the image corresponding to the feature is returned. If the traffic monitoring object is a vehicle, fuzzy recognition of license plate numbers is supported, and users can set the recognition threshold of license plate numbers.

[0090] It is understandable that the traffic monitoring database includes multiple sample traffic driving images, and the administrator can upload images to the traffic monitoring database. That is, this embodiment can also include: obtaining a sample traffic driving image, and calculating the image acquisition quality score of the sample traffic driving; if the image acquisition quality score of the sample traffic driving is greater than or equal to a first preset value, then calculating the similarity between the sample traffic driving image and all images in the traffic monitoring database; when all the similarities are less than a second preset value, saving the sample traffic driving image to the control image library.

[0091] In one implementation, after the urban traffic monitoring center platform receives the sample traffic driving images, it is necessary to judge the image quality, so that images with too low acquisition quality can be filtered out to prevent interference with urban traffic monitoring warnings. For images that pass the quality score screening, it is also necessary to calculate the similarity with the features of existing images in the traffic monitoring database, filter out images with too high similarity, and prevent duplicate sample traffic driving images.

[0092] The urban traffic monitoring and early warning method provided in the embodiment of the present application realizes real-time capture and monitoring and early warning of vehicles or pedestrians. Traffic management agencies can deploy a list of traffic monitoring cameras that need to be monitored and set monitoring and early warning rules to meet flexible monitoring and early warning needs. At the same time, based on the driving plan and image acquisition quality information of vehicles or pedestrians in each traffic driving image, an early warning verification image is determined in the driving video data, ensuring the accuracy of traffic monitoring and early warning.

[0093] This embodiment will be described in detail, specifically:

[0094] See also Figure 3 , Figure 3 A flowchart of another method for updating a traffic monitoring image database provided in an embodiment of the present application,

[0095] S201: Acquire driving video data of a vehicle or a pedestrian; wherein the driving video data includes a plurality of traffic driving images, each of which is an image obtained after image acquisition of the target object;

[0096] The execution subject of this embodiment is Figure 1 The urban traffic monitoring center platform 200 is used to capture vehicles or people in real time.

[0097] S202: Determine the traffic driving image with the highest quality score in the driving video data as a sample reference image, and obtain snapshot information of the sample reference image; wherein the snapshot information at least includes a snapshot location and a snapshot time;

[0098] In this step, the traffic image with the highest image acquisition quality score is selected as the sample reference image, and the captured information of the vehicle or person is extracted from the sample reference image, which may include the capture location, capture time, etc. If the vehicle or person in the sample reference image is a vehicle, the captured information may also include vehicle information. Among them, the vehicle information includes a license plate thumbnail, license plate number, body color, vehicle type, vehicle brand, vehicle system, etc. The license plate thumbnail is specifically the area corresponding to the license plate number in the sample reference image.

[0099] In a specific implementation, the features of the vehicle area can be extracted through a deep learning network, and the vehicle area can be divided into a part with a license plate and a part without a license plate, and finally a license plate thumbnail can be obtained. The features of the vehicle area can be processed. During the specific processing, the vehicle shape, color, pattern, text distribution and other attribute information can be determined through comprehensive extraction to obtain multiple features of the pixels in the determination area, and the license plate number, body color, vehicle type, vehicle brand, and vehicle system can be identified based on the extracted features.

[0100] Since this step extracts the captured information of vehicles or people, traffic monitoring and warning can be performed on any captured information during monitoring and warning, that is, monitoring and warning are preset, and an alarm is triggered when the captured information meets the monitoring and warning. For example, the monitoring and warning rule is that the color of the vehicle is an off-road vehicle. During the real-time capture process, if the vehicle model in the captured information is an off-road vehicle, an alarm is triggered.

[0101] S203: extracting the area corresponding to the vehicle or person from the sample reference image as an image to be stored in a database, and saving the image and the snapshot information corresponding to each image into the database.

[0102] In this step, the area corresponding to the vehicle or person in the image to be extracted is taken as a captured image, and is saved together with the captured information extracted in the previous step in the captured database (traffic monitoring image database) to realize the real-time capture function. As a preferred implementation, this embodiment also includes: displaying all the captured images in the captured database and the captured information corresponding to each of the captured images. In a specific implementation, the captured images and captured information can be displayed in real time on the central screen on the client side to facilitate real-time monitoring.

[0103] The above method can establish a high-quality traffic monitoring image database, which can improve the accuracy and efficiency of monitoring and warning in actual urban traffic monitoring and warning.

[0104] It is understandable that the traffic monitoring image database in this embodiment supports basic management operations, and users can manage it through the client's interactive interface, such as adding images, deleting images, refreshing the database, classifying images, uploading or modifying image information, etc.

[0105] The following is an introduction to an urban traffic monitoring and early warning device based on artificial intelligence provided in an embodiment of the present application. The urban traffic monitoring and early warning device described below and the urban traffic monitoring and early warning method described above can be referenced to each other.

[0106] See also Figure 4 , a structural diagram of a city traffic monitoring and early warning device provided in an embodiment of the present application, such as Figure 4 As shown, it is applicable to the urban traffic monitoring center platform, including:

[0107] Configuration module 301, used to perform monitoring and early warning configuration processing on the urban traffic monitoring task, and obtain monitoring and early warning rules for the urban traffic monitoring task;

[0108] An acquisition module 302 is used to acquire a traffic camera identification set for urban traffic control, wherein the traffic camera identification set indicates a plurality of urban traffic monitoring cameras, and the urban traffic monitoring cameras are deployed at key traffic nodes in the city and are used to collect images of vehicles and pedestrians;

[0109] A collection module 303 is used for collecting driving video data of vehicles or people passing through the traffic node area through the city traffic monitoring camera in response to the execution of the city traffic monitoring task, wherein the driving video data includes at least two traffic driving images of the vehicle or the person;

[0110] The feature processing module 304 is used to perform trajectory feature processing on each traffic driving image of the vehicle or person to obtain the traffic driving trajectory feature of each traffic driving image of the vehicle or person;

[0111] The warning processing module 305 is used to determine the warning verification image of the area corresponding to the traffic node based on the traffic driving trajectory characteristics and the image acquisition quality information of the vehicle or person in each traffic driving image, and the image acquisition quality information is used to measure the quality of the vehicle or person in the traffic driving image collected by the traffic monitoring camera; if the warning verification image of the vehicle or person in the area corresponding to the traffic node meets the monitoring warning rules, a warning notification information is sent to the warning terminal.

[0112] Based on the above embodiment, as a preferred implementation, the warning processing module 305 is specifically used to: generate the traffic driving trajectory of the vehicle or person in the area corresponding to the traffic node according to the traffic driving trajectory characteristics;

[0113] A warning verification image of the area corresponding to the traffic node is determined according to the traffic driving trajectory and the image acquisition quality information.

[0114] Based on the above embodiment, as a preferred implementation, the warning processing module 305 is specifically used for:

[0115] Extracting the traffic driving image to be verified from the driving video data according to the video time sequence;

[0116] If the traffic driving trajectory is within a first preset trajectory range and the image acquisition quality score is greater than or equal to a first preset threshold, the traffic driving image to be verified is determined as a warning verification image, wherein the first preset trajectory range indicates that the vehicle or person is in a normal driving state in the traffic node area;

[0117] If the traffic driving trajectory is within the first preset trajectory range and the image acquisition quality score is less than the first preset threshold, the step of extracting the traffic driving image to be verified from the driving video data in video time order is re-executed; if the traffic driving trajectory is within the second preset trajectory range, the traffic driving image with the highest image acquisition quality score is selected as the early warning verification image, wherein the second preset trajectory range represents that the vehicle or person is in an abnormal driving state in the traffic node area.

[0118] Based on the above embodiment, as a preferred implementation, the warning processing module 305 is specifically used to: if the traffic driving trajectory is within the second preset trajectory range, obtain the target traffic driving image with the highest image acquisition quality score;

[0119] Combining the image acquisition quality score of the target traffic driving image and the multiple traffic driving trajectory features corresponding to the traffic driving trajectory into an image quality index sequence, wherein the image quality index sequence includes at least one image quality index, wherein one image quality index corresponds to one image acquisition quality score or one traffic driving trajectory feature;

[0120] Calculating the standard deviation of each image quality indicator in the image quality indicator sequence to obtain the indicator standard deviation corresponding to each image quality indicator in the image quality indicator sequence, wherein the indicator standard deviation measures the degree of dispersion of the indicators in the image quality indicator sequence;

[0121] Calculating the correlation coefficient of the image quality index sequence to obtain the index correlation coefficient between the image quality indexes, wherein the index correlation coefficient is used to measure the linear relationship between the image quality indexes;

[0122] Determine a weight corresponding to each image quality indicator in the image quality indicator sequence according to an indicator standard deviation corresponding to each image quality indicator and an indicator correlation coefficient between the image quality indicators;

[0123] Performing weighted processing on the image quality indicator sequence according to the weight corresponding to each image quality indicator in the image quality indicator sequence to obtain a comprehensive quality score of the target traffic driving image;

[0124] The target traffic driving image is determined as a warning verification image according to the comprehensive quality score.

[0125] On the basis of the above-mentioned embodiment, as a preferred implementation mode, the warning processing module 305 is specifically used for: matching the warning verification image with the image in the traffic monitoring image database, and if the matching result meets the preset warning condition, sending the warning notification information to the warning terminal.

[0126] On the basis of the above-mentioned embodiment, as a preferred implementation mode, the warning processing module 305 is specifically used for: if the matching result indicates a successful match, querying the area level corresponding to the traffic node in the traffic monitoring database, and determining the traffic warning level of the vehicle or person according to the area level; sending warning notification information corresponding to the traffic warning level to the warning terminal.

[0127] On the basis of the above embodiments, as a preferred implementation mode, the feature processing module 304 is specifically used to preprocess each traffic driving image of the vehicle or person, and use a deep learning model to perform trajectory feature processing on each traffic driving image of the vehicle or person after preprocessing, so as to obtain the traffic driving trajectory features of each traffic driving image of the vehicle or person.

[0128] On the basis of the above embodiments, as a preferred implementation mode, the urban traffic monitoring and early warning device also includes: a quality calculation module, which is specifically used to: obtain image clarity parameters, image position information, image brightness information, image proportion information, and camera angle parameters when the traffic monitoring camera captures images of vehicles or people in traffic; calculate the image acquisition quality score according to the image clarity parameters, image position information, image brightness information, image proportion information, and camera angle parameters.

[0129] As can be seen from the above, the present application provides an urban traffic monitoring and early warning device based on artificial intelligence, which can screen out high-quality early warning verification images from the traffic driving images of vehicles or people captured by the camera, and perform urban traffic early warning based on the early warning verification images, which can improve the real-time performance of urban traffic monitoring and early warning. In addition, since the scheme can select high-quality early warning verification images based on traffic driving trajectory characteristics and image acquisition quality information, the accuracy of urban traffic monitoring and early warning can be improved. Regarding the device in the above embodiment, the specific manner in which each module performs the operation has been described in detail in the embodiment of the method, and will not be elaborated here.

[0130] This application also provides a server, see Figure 5 , a structural diagram of a server 50 provided in an embodiment of the present application, such as Figure 5 As shown, a processor 21 and a memory 22 may be included.

[0131] Among them, the processor 21 may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor 21 can be implemented in at least one hardware form of DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), and PLA (Programmable Logic Array). The processor 21 may also include a main processor and a coprocessor. The main processor is a processor for processing data in the awake state, also known as a CPU (Central Processing Unit); the coprocessor is a low-power processor for processing data in the standby state. In some embodiments, the processor 21 may be integrated with a GPU (Graphics Processing Unit), and the GPU is responsible for rendering and drawing the content to be displayed on the display screen.

[0132] In some embodiments, the processor 21 may also be an AI (Artificial Intelligence) processor, which is used to process computing operations related to machine learning.

[0133] The memory 22 may include one or more computer-readable storage media, which may be non-transitory. The memory 22 may also include a high-speed random access memory, and a non-volatile memory, such as one or more disk storage devices, flash memory storage devices. In this embodiment, the memory 22 is at least used to store the following computer program, wherein, after the computer program is loaded and executed by the processor 21, it can implement the relevant steps in the urban traffic monitoring and early warning method performed by the server side disclosed in any of the aforementioned embodiments. In addition, the resources stored in the memory 22 may also include an operating system and data, etc., and the storage method may be temporary storage or permanent storage. Among them, the operating system may include Windows, Unix, Linux, etc.

[0134] certainly, Figure 5 The server structure shown does not constitute a limitation on the server in the embodiment of the present application. In actual applications, the server may include Figure 5 More or fewer components than shown, or combinations of certain components.

[0135] In another exemplary embodiment, a computer-readable storage medium including program instructions is also provided. When the program instructions are executed by a processor, the steps of the urban traffic monitoring and early warning method executed by the server in any of the above embodiments are implemented.

[0136] The various embodiments in the specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments, and the same and similar parts between the various embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description. It should be pointed out that for ordinary technicians in this technical field, without departing from the principles of this application, several improvements and modifications can be made to the present application, and these improvements and modifications also fall within the scope of protection of the claims of this application.

[0137] It should also be noted that, in this specification, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the statement "comprises a ..." does not exclude the presence of other identical elements in the process, method, article or device including the element.

Claims

1. An urban traffic monitoring and early warning method based on artificial intelligence, characterized in that: Applicable to the urban traffic monitoring center platform, the urban traffic monitoring early warning method includes: Perform monitoring and early warning configuration processing on the urban traffic monitoring task to obtain the monitoring and early warning rules of the urban traffic monitoring task; Obtaining a traffic camera identification set for urban traffic control, wherein the traffic camera identification set indicates a plurality of urban traffic monitoring cameras, and the urban traffic monitoring cameras are deployed at key traffic nodes in the city and are used to collect images of vehicles and pedestrians; In response to the execution of the urban traffic monitoring task, the urban traffic monitoring camera collects driving video data of vehicles or people passing through the traffic node area, wherein the driving video data includes at least two traffic driving images of the vehicles or people; Performing trajectory feature processing on each traffic driving image of the vehicle or person to obtain traffic driving trajectory features of each traffic driving image of the vehicle or person, wherein the traffic driving trajectory features of the vehicle or person refer to a series of descriptive parameters extracted by analyzing the movement path of the vehicle or pedestrian in a certain time and space in the urban traffic monitoring system; Generate a traffic driving trajectory of the vehicle or person in the area corresponding to the traffic node according to the traffic driving trajectory characteristics; obtain image acquisition quality information of the vehicle or person in each traffic driving image; the image acquisition quality information includes an image acquisition quality score, which is used to measure the quality of the vehicle or person in the traffic driving image acquired by the traffic monitoring camera; Extracting the traffic driving image to be verified from the driving video data according to the video time sequence; If the traffic driving trajectory is within a first preset trajectory range and the image acquisition quality score is greater than or equal to a first preset threshold, the traffic driving image to be verified is determined as a warning verification image, wherein the first preset trajectory range indicates that the vehicle or person is in a normal driving state in the traffic node area; If the traffic driving trajectory is within the first preset trajectory range and the image acquisition quality score is less than the first preset threshold, re-performing the step of extracting the traffic driving image to be verified from the driving video data in the video time sequence; If the traffic driving trajectory is within the second preset trajectory range, a target traffic driving image with the highest image acquisition quality score is obtained; wherein the second preset trajectory range indicates that a vehicle or a person is in an abnormal driving state in a traffic node area, then the image acquisition quality score of the target traffic driving image and a plurality of traffic driving trajectory features corresponding to the traffic driving trajectory are combined into an image quality index sequence, wherein the image quality index sequence includes at least one image quality index, wherein one image quality index corresponds to one image acquisition quality score or traffic driving trajectory feature; Calculating the standard deviation of each image quality indicator in the image quality indicator sequence to obtain the indicator standard deviation corresponding to each image quality indicator in the image quality indicator sequence, wherein the indicator standard deviation measures the degree of dispersion of the indicators in the image quality indicator sequence; Calculating the correlation coefficient of the image quality index sequence to obtain the index correlation coefficient between the image quality indexes, wherein the index correlation coefficient is used to measure the linear relationship between the image quality indexes; Determine a weight corresponding to each image quality indicator in the image quality indicator sequence according to an indicator standard deviation corresponding to each image quality indicator and an indicator correlation coefficient between the image quality indicators; Performing weighted processing on the image quality indicator sequence according to the weight corresponding to each image quality indicator in the image quality indicator sequence to obtain a comprehensive quality score of the target traffic driving image; Determining the target traffic driving image as a warning verification image according to the comprehensive quality score; wherein if the comprehensive quality score of the target traffic driving image is greater than a preset threshold, determining the target traffic driving image as a warning verification image, otherwise, it is necessary to select the remaining traffic driving images as warning verification images; If the warning verification image of the vehicle or person in the area corresponding to the traffic node meets the monitoring warning rule, a warning notification message is sent to the warning terminal.

2. The urban traffic monitoring and early warning method according to claim 1 is characterized in that: If the warning verification image of the vehicle or person in the area corresponding to the traffic node meets the monitoring warning rule, a warning notification message is sent to the warning terminal, including: The warning verification image is matched with the image in the traffic monitoring image database, and if the matching result meets the preset warning condition, a warning notification information is sent to the warning terminal.

3. The urban traffic monitoring and early warning method according to claim 1 is characterized in that: If the warning verification image of the vehicle or person in the area corresponding to the traffic node meets the monitoring warning rule, a warning notification message is sent to the warning terminal, including: If the matching result indicates a successful match, querying the regional level corresponding to the traffic node in the traffic monitoring database, and determining the traffic warning level of the vehicle or person according to the regional level; Sending warning notification information corresponding to the traffic warning level to the warning terminal.

4. The urban traffic monitoring and early warning method according to claim 3 is characterized in that: Performing trajectory feature processing on each traffic driving image of the vehicle or person to obtain traffic driving trajectory features of each traffic driving image of the vehicle or person includes: Preprocessing is performed on each traffic driving image of the vehicle or person, and trajectory feature processing is performed on each traffic driving image of the vehicle or person after the preprocessing using a deep learning model to obtain traffic driving trajectory features of each traffic driving image of the vehicle or person.

5. The urban traffic monitoring and early warning method according to claim 4 is characterized in that: The method further comprises: Obtain image clarity parameters, image position information, image brightness information, image proportion information, and camera angle parameters when a traffic monitoring camera collects images of vehicles or people in traffic; The image acquisition quality score is calculated based on image clarity parameters, image position information, image brightness information, image proportion information, and camera angle parameters.

6. An urban traffic monitoring and early warning device based on artificial intelligence, characterized in that: Applicable to the city traffic monitoring center platform, the city traffic monitoring and early warning device includes: The configuration module is used to perform monitoring and early warning configuration processing on the urban traffic monitoring task and obtain the monitoring and early warning rules of the urban traffic monitoring task; An acquisition module is used to acquire a traffic camera identification set for urban traffic control, wherein the traffic camera identification set indicates a plurality of urban traffic monitoring cameras, and the urban traffic monitoring cameras are deployed at key traffic nodes in the city and are used to collect images of vehicles and pedestrians; A collection module, configured to collect driving video data of vehicles or people passing through the traffic node area through the city traffic monitoring camera in response to the execution of the city traffic monitoring task, wherein the driving video data includes at least two traffic driving images of the vehicle or the person; A feature processing module is used to perform trajectory feature processing on each traffic driving image of the vehicle or person to obtain the traffic driving trajectory feature of each traffic driving image of the vehicle or person, wherein the traffic driving trajectory feature of the vehicle or person refers to a series of descriptive parameters extracted by analyzing the movement path of the vehicle or pedestrian in a certain time and space in the urban traffic monitoring system; An early warning processing module is used to generate a traffic driving trajectory of the vehicle or person in the corresponding area of ​​the traffic node according to the traffic driving trajectory characteristics; obtain image acquisition quality information of the vehicle or person in each traffic driving image; the image acquisition quality information includes an image acquisition quality score, which is used to measure the quality of the vehicle or person in the traffic driving image acquired by the traffic monitoring camera; extract the traffic driving image to be verified from the driving video data in the video time sequence; if the traffic driving trajectory is within a first preset trajectory range and the image acquisition quality score is greater than or equal to a first preset threshold, then determine the traffic driving image to be verified as a warning verification image, wherein the first preset trajectory range The method is to characterize that a vehicle or a person is in a normal driving state in a traffic node area; if the traffic driving trajectory is within a first preset trajectory range and the image acquisition quality score is less than a first preset threshold, re-execute the step of extracting the traffic driving image to be verified from the driving video data in the video time sequence; if the traffic driving trajectory is within a second preset trajectory range, obtain the target traffic driving image with the highest image acquisition quality score; wherein the second preset trajectory range characterizes that a vehicle or a person is in an abnormal driving state in a traffic node area, then the image acquisition quality score of the target traffic driving image and the multiple traffic driving trajectory features corresponding to the traffic driving trajectory are combined into an image quality indicator sequence, and the image quality indicator sequence The method comprises at least one image quality indicator, wherein one image quality indicator corresponds to an image acquisition quality score or a traffic driving trajectory feature; performing standard deviation calculation on each image quality indicator in the image quality indicator sequence to obtain the indicator standard deviation corresponding to each image quality indicator in the image quality indicator sequence, wherein the indicator standard deviation measures the degree of dispersion of the indicator in the image quality indicator sequence; performing correlation coefficient calculation on the image quality indicator sequence to obtain the indicator correlation coefficient between each image quality indicator, wherein the indicator correlation coefficient is used to measure the linear relationship between the image quality indicators; determining the linear relationship between the image quality indicators according to the indicator standard deviation corresponding to each image quality indicator and the indicator correlation coefficient between the image quality indicators. Determine the weight corresponding to each image quality indicator in the image quality indicator sequence; perform weighted processing on the image quality indicator sequence according to the weight corresponding to each image quality indicator in the image quality indicator sequence to obtain the comprehensive quality score of the target traffic driving image; determine the target traffic driving image as a warning verification image according to the comprehensive quality score; if the comprehensive quality score of the target traffic driving image is greater than a preset threshold, determine the target traffic driving image as a warning verification image, otherwise, it is necessary to select the remaining traffic driving images as warning verification images; if the warning verification image of the vehicle or person in the corresponding area of ​​the traffic node meets the monitoring warning rule, send a warning notification information to the warning terminal.

7. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the urban traffic monitoring and early warning method as claimed in any one of claims 1 to 5.

Images