Traffic flow management and control system and method based on visual inspection

Abstract

The invention relates to the field of traffic flow management and control based on visual inspection, in particular to a traffic flow management and control system and method based on visual inspection, and the system comprises an image collection module, a traffic control module, a vehicle recognition module, an image analysis module, an information issuing module and a remote monitoring module. The traffic control module automatically adjusts the duration of the signal lamp according to the real-time position and the driving direction of the vehicle, and the image analysis module analyzes the number and the speed of the vehicle according to the collected driving video of each vehicle and counts the traffic flow of different road sections and time periods. And vehicle information and traffic flow conditions are identified, so that intelligent control of traffic signals is realized, real-time traffic information is published to the public, and traffic efficiency and safety are improved.

Description

Technical Field

[0001] The present invention relates to the field of traffic flow control based on visual detection, and specifically to a traffic flow control system and method based on visual detection. Background Art

[0002] With the acceleration of urbanization and the continuous growth of car ownership, traffic congestion has become a serious problem faced by major cities around the world. Traffic congestion not only leads to longer travel time, energy waste and environmental pollution, but also affects the economic development of the city and the quality of life of residents. Traditional traffic flow control methods mainly rely on manual monitoring and experience judgment, which is difficult to meet the needs of modern traffic management. Therefore, a more efficient, accurate and intelligent traffic flow control system is needed to solve the problem of traffic congestion.

[0003] In recent years, vision technology has made great progress and achieved remarkable results in the fields of target detection, image recognition, video analysis, etc. Vision technology can collect image and video data through cameras and other devices, and then use image processing and pattern recognition algorithms to analyze and process the data to extract valuable information. Intelligent transportation systems can realize the collection, transmission, processing and release of traffic information, provide timely and accurate traffic information services for traffic management departments and travelers, and improve the operating efficiency and safety of transportation systems. The traffic flow control system based on visual detection is an important part of the intelligent transportation system, which can provide real-time and accurate traffic flow data for the intelligent transportation system and provide support for traffic management and decision-making. Summary of the invention

[0004] In order to solve the technical problems raised by the above-mentioned background technology, the present invention provides a traffic flow control system and method based on visual detection.

[0005] The purpose of the present invention can be achieved by the following technical solutions: A first aspect of the present invention provides a traffic flow control system based on visual detection, including an image acquisition module, a vehicle recognition module, an image analysis module, a traffic control module, an information release module, a remote monitoring module, a data center and a cloud server.

[0006] The image acquisition module collects images of each vehicle using a high-definition camera. The specific acquisition process is as follows: The output end of the image acquisition module is connected to the input end of the image analysis module. Multiple high-definition cameras are installed at key locations of traffic arteries, including indicator lights in all directions of intersections, entrances and exits of highways, and upper and lower bridges of viaducts. The high-definition cameras are equipped with automatic dimming and infrared night vision functions. The high-definition cameras are used to collect driving videos of various types of vehicles in real time, and the driving videos of various types of vehicles are obtained. The driving videos of various vehicles are pre-processed, including denoising, enhancement, and format conversion, to obtain pre-processed videos of various vehicles, and the pre-processed videos of various vehicles are sent to the image analysis module. The vehicle identification module obtains the information images of each vehicle in real time according to each high-definition camera, and identifies the vehicle according to the information images. The specific acquisition process is as follows: The input end of the vehicle identification module is connected to the output end of the cloud server. The vehicle identification module identifies each vehicle type based on shape features. Each vehicle type includes cars, buses and trucks. Specifically, the actual contour shape of each type of vehicle is obtained through each high-definition camera. The actual contour shape includes actual length, actual width and overall shape. The actual contour shape corresponding to each type of vehicle is obtained. The preset length and preset width corresponding to each type of vehicle are obtained through the cloud server. The actual length and actual width of each type of vehicle are matched with the preset length and preset width of the corresponding vehicle respectively. If the match is successful, the corresponding vehicle is marked as a compliant vehicle, and then the vehicle type is determined based on the overall shape. If the overall shape of the vehicle is relatively round, it is classified as a car type. If the overall shape of the vehicle is relatively square, it is classified as a truck type. If the overall shape of the vehicle is very long, it is classified as a bus type. In this way, the current vehicle types are obtained.

[0007] Special vehicles are identified based on color features, specifically: various types of vehicle images are obtained through high-definition cameras, and color information of various types of vehicle images is extracted. If the body color of the corresponding vehicle is white with red and blue warning stripes, and there are signs such as "First Aid" on the body, it will be marked as an emergency vehicle. If the body color of the corresponding vehicle is red with fire signs and large fire-fighting equipment, it will be marked as a fire truck. When the high-definition camera terminal identifies a fire truck and an emergency vehicle, it generates a warning signal and sends it to the traffic signal control module.

[0008] The output end of the vehicle identification module is connected to the input end of the data center, and the license plate is identified based on digital features and color features. Specifically, various types of vehicle images are obtained through a high-definition camera, and the license plate numbers of various types of vehicle images are identified using optical character recognition technology. The license plate image is projected in the vertical direction, and the boundaries of the characters are determined according to the peaks and troughs of the projection curve. In the projection curve, the peaks correspond to the spacing area between the characters, and the troughs correspond to the positions of the characters. The characters on the license plate are segmented into individual character images according to the determined character boundaries, and a template containing various characters is established in the data center. The segmented character images are matched one by one with the characters in the template database to obtain each matching character, and the corresponding vehicle information is obtained through each matching character.

[0009] The image analysis module analyzes the collected driving videos of each vehicle, and the specific acquisition process is as follows: Receive the pre-processed videos of each vehicle sent by the image acquisition module, convert the pre-processed videos of each vehicle into driving pictures of each vehicle frame by frame, obtain the driving pictures of each vehicle, obtain the corresponding picture pixels according to the pictures of each type of vehicle, calculate the difference between the corresponding picture pixels of each type of vehicle pictures in two adjacent frames, extract the preset picture pixel difference value of the data center, compare the pixel difference value of the actual picture with the preset picture pixel difference value, if the pixel difference value of the actual picture is less than the preset picture pixel difference value, it means that the vehicle has not moved, on the contrary, it means that each type of vehicle has moved at this position, and obtain the conversion value of pixel distance and actual physical distance through camera calibration, and obtain the distance generated in the picture through the conversion value, marked as DE, and calculated using the formula Get the corresponding vehicle instantaneous speed NH, CV represents the time interval between two frames, extract the preset instantaneous speed corresponding to the road section in the data center, compare the actual instantaneous speed with the preset instantaneous speed corresponding to the road section, if the actual instantaneous speed is greater than the preset instantaneous speed corresponding to the road section, it means that the vehicle speed exceeds the prescribed speed.

[0010] A day is divided into multiple time periods, including the morning peak period from 7:00 to 9:00, noon from 11:00 to 13:00, evening peak from 17:00 to 19:00, and night from 21:00 to 5:00 in the morning. Traffic flow statistics are then performed on different roads in the city, including urban expressways, main roads, secondary roads, and branch roads. Each road section is divided into several small sections, and the length of the small section is obtained, marked as BN. The vehicles passing through several small sections in different time periods are counted by high-definition cameras to obtain the number of vehicles in each section in each time period, marked as LP. The formula is used The vehicle density GH of the corresponding road section is obtained. The driving images of each type of vehicle corresponding to each small road section in each time period are collected by high-definition cameras. The corresponding driving images of each type of vehicle are obtained. The driving speed of each vehicle is obtained through the driving images of each vehicle. The driving speed of each vehicle is averaged to obtain the average driving speed of each vehicle, marked as HF. The calculation is performed by the number of vehicles, vehicle density and average vehicle speed, and the formula is used. The corresponding road section congestion value DS is obtained, where G1, G2 and G3 are respectively represented as preset proportional coefficients, the preset road section congestion value of the data center is extracted, and the road section congestion value is compared with the preset road section congestion value. If the actual road section congestion value is greater than the preset congestion value, it means that there are too many vehicles in the current road section, causing congestion, and a clearing signal is generated and sent to the traffic signal control module.

[0011] The traffic control module receives various types of signals and processes them separately. The specific acquisition methods are as follows: The output end of the vehicle identification module is connected to the input end of the traffic flow control module. By receiving the warning signal sent by the vehicle identification module, the high-definition camera identifies the location information of the corresponding special vehicle and sends the corresponding location information to the traffic flow control terminal. The traffic flow control terminal receives the location information of the corresponding special vehicle, generates a control instruction and sends it to the corresponding traffic light processing terminal. The traffic light processing terminal controls the traffic light according to the received control instruction. If the current traffic light is red, it is adjusted to green. If the current traffic light is green, the green light passage time is extended until the special vehicle drives out of the monitoring range of the high-definition camera. The traffic flow control generates a stop instruction and sends it to the traffic light processing terminal, and the traffic light stops the current operation.

[0012] After receiving the unblocking signal sent by the image analysis module, the traffic control module sends the unblocking signal to the traffic flow control terminal. The traffic flow control terminal obtains the corresponding congestion location according to the high-definition camera, takes the congestion location as the origin, searches for the traffic control personnel with the smallest distance within the radius through the mobile terminal signal, generates an adjustment instruction and sends it to the mobile terminal of the corresponding traffic control personnel. The traffic control personnel receives the adjustment instruction and Order, go to the corresponding road section to clear the blockage.

[0013] The information release module releases the information through the electronic display screen according to the congestion and emergency conditions of each road section. The specific acquisition process is as follows: The output end of the information release module is connected to the input end of the cloud server, and the electronic display screens are distributed in conspicuous locations such as both sides of the road, intersections and highway service areas. The congestion conditions, accident information and road construction information of each road section are obtained in real time through high-definition cameras, and the obtained road information is sent to the electronic display screen terminal in real time. The electronic display screen terminal receives the road information and plays it through the display screen.

[0014] The remote monitoring module uses high-definition cameras to monitor the real-time conditions of each road section and solve emergencies. The specific acquisition process is as follows: A large-screen display system is set up in the remote monitoring center to display the video images and traffic data of each monitoring point in real time. The large-screen display system has multi-screen switching, zooming and playback functions. Each traffic control personnel monitors each road section in real time through the large-screen display system. When any unexpected situation is found, the nearby traffic control personnel are contacted in real time through the traffic flow control terminal for processing.

[0015] The second aspect of the present invention provides a traffic flow control method based on visual detection, and the specific steps are as follows: Step 1: Image acquisition: collect videos of each vehicle through a high-definition camera and send the collected videos to image analysis; Step 2: Vehicle identification: Use high-definition cameras to identify the characteristics of various special vehicles and types of vehicles in real time; Step 3: Image analysis: Analyze the number and speed of vehicles based on the received video information, and count the traffic flow in different sections and time periods; Step 4: Traffic control: Automatically adjust the phase and duration of traffic lights according to the real-time location and driving direction of the vehicle to ensure the orderly passage of traffic flow; Step 5: Information release: Based on the congestion and emergency situations of each road section, information is released through the electronic display screen; Step 6. Remote monitoring: Use high-definition cameras to monitor the real-time conditions of each road section and resolve emergencies.

[0016] Compared with the prior art, the beneficial effects of the present invention are: improving traffic efficiency, through visual detection equipment, vehicles on the road can be continuously monitored, and traffic flow conditions in different time periods and different sections can be accurately grasped. This allows the traffic management department to timely adjust the traffic signal timing and optimize the allocation of road resources according to the actual traffic flow, avoid congestion caused by excessive traffic flow, and improve road traffic capacity. During peak hours, if the system detects a sharp increase in traffic flow on a certain main road, it can automatically extend the green light time of the section, reduce vehicle waiting time, and thus speed up vehicle traffic. The system based on visual detection can dynamically adjust traffic signals according to real-time traffic conditions. When it is detected that there is a long queue of vehicles in a certain direction, the system can immediately adjust the signal light to increase the green light time in that direction, thereby reducing vehicle congestion and improving the traffic efficiency of the intersection.

[0017] The visual inspection system can quickly identify traffic anomalies such as traffic accidents and vehicle breakdowns. Once an anomaly is detected, the system can immediately notify the traffic management department so that rescue and diversion measures can be taken in time to reduce the occurrence of secondary accidents. When the system detects a traffic accident on a certain road section, it can automatically send instructions to nearby traffic lights, adjust the status of the lights, guide vehicles to detour, and notify the traffic police department to handle the accident, thereby reducing the impact of the accident on traffic and ensuring road traffic safety. The traffic management department can reasonably arrange the deployment of police forces based on the real-time traffic information provided by the system, increase police forces in areas with traffic congestion or frequent accidents, and conduct timely diversion and handling to improve the efficiency of traffic management. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. The following drawings are not intentionally scaled to the actual size, and the focus is on illustrating the main purpose of the present invention.

[0019] Figure 1 It is a principle block diagram of the present invention.

[0020] Figure 2 These are the method steps of the present invention. DETAILED DESCRIPTION

[0021] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work also fall within the scope of protection of the present invention.

[0022] Example 1: Figure 1 As shown, the present invention provides a traffic flow control system based on visual detection, including an image acquisition module, a vehicle recognition module, an image analysis module, a traffic control module, an information release module, a remote monitoring module, a data center and a cloud server.

[0023] The image acquisition module collects images of each vehicle using a high-definition camera. The specific acquisition process is as follows: The output end of the image acquisition module is connected to the input end of the image analysis module. Multiple high-definition cameras are installed at key locations of traffic arteries, including indicator lights in all directions of intersections, entrances and exits of highways, and upper and lower bridges of viaducts. The high-definition cameras are equipped with automatic dimming and infrared night vision functions. The high-definition cameras are used to collect driving videos of various types of vehicles in real time, and the driving videos of various types of vehicles are obtained. The driving videos of various vehicles are pre-processed, including denoising, enhancement, and format conversion, to obtain pre-processed videos of various vehicles, and the pre-processed videos of various vehicles are sent to the image analysis module. The vehicle identification module obtains the information images of each vehicle in real time according to each high-definition camera, and identifies the vehicle according to the information images. The specific acquisition process is as follows: The input end of the vehicle identification module is connected to the output end of the cloud server. The vehicle identification module identifies each vehicle type based on shape features. Each vehicle type includes cars, buses and trucks. Specifically, the actual contour shape of each type of vehicle is obtained through each high-definition camera. The actual contour shape includes actual length, actual width and overall shape. The actual contour shape corresponding to each type of vehicle is obtained. The preset length and preset width corresponding to each type of vehicle are obtained through the cloud server. The actual length and actual width of each type of vehicle are matched with the preset length and preset width of the corresponding vehicle respectively. If the match is successful, the corresponding vehicle is marked as a compliant vehicle, and then the vehicle type is determined based on the overall shape. If the overall shape of the vehicle is relatively round, it is classified as a car type. If the overall shape of the vehicle is relatively square, it is classified as a truck type. If the overall shape of the vehicle is very long, it is classified as a bus type. In this way, the current vehicle types are obtained.

[0024] Special vehicles are identified based on color features, specifically: various types of vehicle images are obtained through high-definition cameras, and color information of various types of vehicle images is extracted. If the body color of the corresponding vehicle is white with red and blue warning stripes, and there are signs such as "First Aid" on the body, it will be marked as an emergency vehicle. If the body color of the corresponding vehicle is red with fire signs and large fire-fighting equipment, it will be marked as a fire truck. When the high-definition camera terminal identifies a fire truck and an emergency vehicle, it generates a warning signal and sends it to the traffic signal control module.

[0025] The output end of the vehicle identification module is connected to the input end of the data center, and the license plate is identified based on digital features and color features. Specifically, various types of vehicle images are obtained through a high-definition camera, and the license plate numbers of various types of vehicle images are identified using optical character recognition technology. The license plate image is projected in the vertical direction, and the boundaries of the characters are determined according to the peaks and troughs of the projection curve. In the projection curve, the peaks correspond to the spacing area between the characters, and the troughs correspond to the positions of the characters. The characters on the license plate are segmented into individual character images according to the determined character boundaries, and a template containing various characters is established in the data center. The segmented character images are matched one by one with the template characters of the data center to obtain each matching character, and the corresponding vehicle information is obtained through each matching character.

[0026] The image analysis module analyzes the collected driving videos of each vehicle, and the specific acquisition process is as follows: Receive the pre-processed videos of each vehicle sent by the image acquisition module, convert the pre-processed videos of each vehicle into driving pictures of each vehicle frame by frame, obtain the driving pictures of each vehicle, obtain the corresponding picture pixels according to the pictures of each type of vehicle, calculate the difference between the corresponding picture pixels of each type of vehicle pictures in two adjacent frames, extract the preset picture pixel difference value of the data center, compare the pixel difference value of the actual picture with the preset picture pixel difference value, if the pixel difference value of the actual picture is less than the preset picture pixel difference value, it means that the vehicle has not moved, on the contrary, it means that each type of vehicle has moved at this position, and obtain the conversion value of pixel distance and actual physical distance through camera calibration, and obtain the distance generated in the picture through the conversion value, marked as DE, and calculated using the formula Get the corresponding vehicle instantaneous speed NH, CV represents the time interval between two frames, extract the preset instantaneous speed corresponding to the road section in the data center, compare the actual instantaneous speed with the preset instantaneous speed corresponding to the road section, if the actual instantaneous speed is greater than the preset instantaneous speed corresponding to the road section, it means that the vehicle speed exceeds the prescribed speed.

[0027] A day is divided into multiple time periods, including the morning peak period from 7:00 to 9:00, noon from 11:00 to 13:00, evening peak from 17:00 to 19:00, and night from 21:00 to 5:00 in the morning. Traffic flow statistics are then performed on different roads in the city, including urban expressways, main roads, secondary roads, and branch roads. Each road section is divided into several small sections, and the length of the small section is obtained, marked as BN. The vehicles passing through several small sections in different time periods are counted by high-definition cameras to obtain the number of vehicles in each section in each time period, marked as LP. The formula is used The vehicle density GH of the corresponding road section is obtained. The driving images of each type of vehicle corresponding to each small road section in each time period are collected by high-definition cameras. The corresponding driving images of each type of vehicle are obtained. The driving speed of each vehicle is obtained through the driving images of each vehicle. The driving speed of each vehicle is averaged to obtain the average driving speed of each vehicle, marked as HF. The calculation is performed by the number of vehicles, vehicle density and average vehicle speed, and the formula is used. The corresponding road section congestion value DS is obtained, where G1, G2 and G3 are respectively represented as preset proportional coefficients, the preset road section congestion value of the data center is extracted, and the road section congestion value is compared with the preset road section congestion value. If the actual road section congestion value is greater than the preset congestion value, it means that there are too many vehicles in the current road section, causing congestion, and a clearing signal is generated and sent to the traffic signal control module.

[0028] The traffic control module receives various types of signals and processes them separately. The specific acquisition methods are as follows: The output end of the vehicle identification module is connected to the input end of the traffic flow control module. By receiving the warning signal sent by the vehicle identification module, the high-definition camera identifies the location information of the corresponding special vehicle and sends the corresponding location information to the traffic flow control terminal. The traffic flow control terminal receives the location information of the corresponding special vehicle, generates a control instruction and sends it to the corresponding traffic light processing terminal. The traffic light processing terminal controls the traffic light according to the received control instruction. If the current traffic light is red, it is adjusted to green. If the current traffic light is green, the green light passage time is extended until the special vehicle drives out of the monitoring range of the high-definition camera. The traffic flow control generates a stop instruction and sends it to the traffic light processing terminal, and the traffic light stops the current operation.

[0029] After receiving the clearing signal sent by the image analysis module, the traffic control module sends the clearing signal to the traffic flow control terminal. The traffic flow control terminal obtains the corresponding congestion location according to the high-definition camera, takes the congestion location as the origin, searches for the traffic control personnel with the smallest distance within the radius through the mobile terminal signal, generates an adjustment instruction and sends it to the mobile terminal of the corresponding traffic control personnel. The traffic control personnel receives the adjustment instruction and goes to the corresponding section to clear the road.

[0030] The information release module releases the information through the electronic display screen according to the congestion and emergency conditions of each road section. The specific acquisition process is as follows: The output end of the information release module is connected to the input end of the cloud server, and the electronic display screens are distributed in conspicuous locations such as both sides of the road, intersections and highway service areas. The congestion conditions, accident information and road construction information of each road section are obtained in real time through high-definition cameras, and the obtained road information is sent to the electronic display screen terminal in real time. The electronic display screen terminal receives the road information and plays it through the display screen.

[0031] The remote monitoring module uses high-definition cameras to monitor the real-time conditions of each road section and solve emergencies. The specific acquisition process is as follows: A large-screen display system is set up in the remote monitoring center to display the video images and traffic data of each monitoring point in real time. The large-screen display system has multi-screen switching, zooming and playback functions. Each traffic control personnel monitors each road section in real time through the large-screen display system. When any unexpected situation is found, the nearby traffic control personnel are contacted in real time through the traffic flow control terminal for processing.

[0032] Example 2: Figure 2 As shown, the present invention provides a traffic flow control method based on visual detection, and the specific steps are as follows: Step 1: Image acquisition: collect videos of each vehicle through a high-definition camera and send the collected videos to image analysis; Step 2: Vehicle identification: Use high-definition cameras to identify the characteristics of various special vehicles and types of vehicles in real time; Step 3: Image analysis: Analyze the number and speed of vehicles based on the received video information, and count the traffic flow in different sections and time periods; Step 4: Traffic control: Automatically adjust the phase and duration of traffic lights according to the real-time location and driving direction of the vehicle to ensure the orderly passage of traffic flow; Step 5: Information release: Based on the congestion and emergency situations of each road section, information is released through the electronic display screen; Step 6. Remote monitoring: Use high-definition cameras to monitor the real-time conditions of each road section and resolve emergencies.

[0033] The above is an explanation of the present invention and should not be considered as a limitation thereof. Although several exemplary embodiments of the present invention have been described, it will be readily appreciated by those skilled in the art that many modifications may be made to the exemplary embodiments without departing from the novel teachings and advantages of the present invention. Therefore, all such modifications are intended to be included within the scope of the present invention as defined in the claims. It should be understood that the above is an explanation of the present invention and should not be considered as being limited to the specific embodiments disclosed, and modifications to the disclosed embodiments and other embodiments are intended to be included within the scope of the appended claims. The present invention is defined by the claims and their equivalents.

Claims

1. A traffic flow control system based on visual detection, comprising an image acquisition module, a vehicle recognition module, an image analysis module, an information release module, a remote monitoring module, a data center and a cloud server, characterized in that: Also included is a traffic control module; The traffic control module automatically adjusts the phase and duration of the traffic light according to the real-time position and driving direction of the vehicle. The output end of the vehicle identification module is connected to the input end of the traffic flow control module. By receiving the warning signal sent by the vehicle identification module, the high-definition camera identifies the position information of the corresponding special vehicle and sends the corresponding position information to the traffic flow control terminal. The traffic flow control terminal receives the position information of the corresponding special vehicle, generates a control instruction and sends it to the corresponding traffic light processing terminal. The traffic light processing terminal controls the traffic light according to the received control instruction. If the current traffic light is red, it is adjusted to green. If the current traffic light is green, the green light passage time is extended until the special vehicle drives out of the monitoring range of the high-definition camera. The traffic flow control generates a stop instruction and sends it to the traffic light processing terminal, and the traffic light stops the current operation. After receiving the clearing signal sent by the image analysis module, the traffic control module sends the clearing signal to the traffic flow control terminal. The traffic flow control terminal obtains the corresponding congestion location based on the high-definition camera, takes the congestion location as the origin, searches for the traffic control personnel with the smallest distance within the radius through the mobile terminal signal, generates an adjustment instruction and sends it to the mobile terminal of the corresponding traffic control personnel. The traffic control personnel receive the adjustment instruction and go to the corresponding section to clear the traffic. The image analysis module analyzes the number and speed of vehicles based on the collected driving videos of each vehicle, and counts the traffic flow in different sections and time periods. The specific process is as follows: Receive the pre-processed videos of each vehicle sent by the image acquisition module, convert the pre-processed videos of each vehicle into driving pictures of each vehicle frame by frame, obtain the driving pictures of each vehicle, obtain the corresponding picture pixels according to the pictures of each type of vehicle, calculate the difference between the corresponding picture pixels of each type of vehicle pictures in two adjacent frames, extract the preset picture pixel difference value of the data center, compare the pixel difference value of the actual picture with the preset picture pixel difference value, if the pixel difference value of the actual picture is less than the preset picture pixel difference value, it means that the vehicle has not moved, on the contrary, it means that each type of vehicle has moved at this position, and obtain the conversion value of pixel distance and actual physical distance through camera calibration, and obtain the distance generated in the picture through the conversion value, marked as DE, and calculated using the formula Get the corresponding vehicle instantaneous speed NH, CV represents the time interval between two frames, extract the preset instantaneous speed corresponding to the road section in the data center, compare the obtained actual instantaneous speed with the preset instantaneous speed corresponding to the road section, if the actual instantaneous speed is greater than the preset instantaneous speed corresponding to the road section, it means that the vehicle speed exceeds the prescribed speed; Divide a day into multiple time periods, and then count the traffic flow on different roads in the city. Divide each road section into several small sections, and then obtain the length of the small section, marked as BN. Use high-definition cameras to count the number of vehicles passing through several small sections in different time periods, and get the number of vehicles in each section in each time period, marked as LP. Use the formula The vehicle density GH of the corresponding road section is obtained. The driving images of each type of vehicle corresponding to each small road section in each time period are collected by high-definition cameras. The corresponding driving images of each type of vehicle are obtained. The driving speed of each vehicle is obtained through the driving images of each vehicle. The driving speed of each vehicle is averaged to obtain the average driving speed of each vehicle, marked as HF. The calculation is performed by the number of vehicles, vehicle density and average vehicle speed, and the formula is used. The corresponding road section congestion value DS is obtained, where G1, G2 and G3 are respectively represented as preset proportional coefficients. The preset road section congestion value of the data center is extracted, and the road section congestion value is compared with the preset road section congestion value. If the actual road section congestion value is greater than the preset congestion value, it means that there are too many vehicles in the current road section, causing congestion, and a clearing signal is generated to the traffic signal control module. Different roads include urban expressways, main roads, secondary roads and branch roads.

2. A traffic flow control system based on visual detection according to claim 1, characterized in that: The information release module releases the information through the electronic display screen according to the congestion and emergency conditions of each road section. The specific process is as follows: The output end of the information release module is connected to the input end of the cloud server, and the electronic display screens are distributed in conspicuous locations on both sides of the road, intersections and highway service areas. The congestion conditions, accident information and road construction information of each road section are obtained in real time through high-definition cameras, and the obtained road information is sent to the electronic display screen terminal in real time. The electronic display screen terminal receives the road information and plays it through the display screen.

3. A traffic flow control system based on visual detection according to claim 1, characterized in that: The remote monitoring module uses high-definition cameras to monitor the real-time conditions of each road section and resolve emergencies. The specific process is as follows: A large-screen display system is set up in the remote monitoring center to display the video images and traffic data of each monitoring point in real time. Traffic control personnel monitor each road section in real time through the large-screen display system. When any unexpected situation is found, they contact nearby traffic control personnel in real time through the traffic flow control terminal for processing.

4. A traffic flow control system based on visual detection according to claim 1, characterized in that: The image acquisition module collects images of each vehicle using a high-definition camera. The specific process is as follows: The output end of the image acquisition module is connected to the input end of the image analysis module, and multiple high-definition cameras are installed at key locations of traffic arteries, including indicator lights in various directions at intersections, entrances and exits of highways, and upper and lower bridges of viaducts. The high-definition cameras are equipped with automatic dimming and infrared night vision functions. The high-definition cameras are used to collect driving videos of various types of vehicles in real time, and the driving videos of various types of vehicles are obtained. The obtained driving videos of each vehicle are pre-processed, and the pre-processing includes denoising, enhancement and format conversion to obtain pre-processed videos of each vehicle, and the pre-processed videos of each vehicle are sent to the image analysis module.

5. The traffic flow control system based on visual detection according to claim 1 is characterized in that: The vehicle identification module obtains the information images of each vehicle in real time through each high-definition camera and identifies the information images of each vehicle. The specific process is as follows: The input end of the vehicle identification module is connected to the output end of the cloud server. The vehicle identification module identifies each vehicle type based on shape features. Each vehicle type includes cars, buses and trucks. Specifically, the actual contour shape of each type of vehicle is obtained through each high-definition camera. The actual contour shape includes actual length, actual width and overall shape. The actual contour shape corresponding to each type of vehicle is obtained. The preset length and preset width corresponding to each type of vehicle are obtained through the cloud server. The actual length and actual width of each type of vehicle are matched with the preset length and preset width of the corresponding vehicle respectively. If the match is successful, the corresponding vehicle is marked as a qualified vehicle. Then the vehicle type is determined based on the overall shape. If the overall shape of the vehicle is relatively round, it is classified as a car. If the overall shape of the vehicle is relatively square, it is classified as a truck. If the overall shape of the vehicle is long, it is classified as a bus. In this way, the current vehicle types are obtained. Identify various special vehicles based on color features, specifically: obtain various types of vehicle images through high-definition cameras, extract color information of various types of vehicle images, and if the corresponding vehicle body color is white with red and blue warning stripes, and there are signs such as "First Aid" on the body, it will be marked as an emergency vehicle. If the corresponding vehicle body is red with fire signs and large fire equipment, it will be marked as a fire truck. When the high-definition camera terminal identifies a fire truck and an emergency vehicle, it generates a warning signal and sends it to the traffic signal control module; The output end of the vehicle identification module is connected to the input end of the data center, and the license plate is identified based on digital features and color features. Specifically, various types of vehicle images are obtained through a high-definition camera, and the license plate numbers of various types of vehicle images are identified using optical character recognition technology. The license plate image is projected in the vertical direction, and the boundaries of the characters are determined according to the peaks and troughs of the projection curve. In the projection curve, the peaks correspond to the spacing area between the characters, and the troughs correspond to the positions of the characters. The characters on the license plate are segmented into individual character images according to the determined character boundaries, and a template containing various characters is established in the data center. The segmented character images are matched one by one with the template characters of the data center to obtain each matching character, and the corresponding vehicle information is obtained through each matching character.

6. A traffic flow control method based on visual detection, characterized in that: A traffic flow control system based on visual detection applied to any one of claims 1 to 5, comprising the following steps: A1. Collect videos of each vehicle through high-definition cameras and send the collected videos to image analysis; A2. Real-time identification of special vehicles and various types of vehicle features through high-definition cameras; A3. Analyze the number and speed of vehicles based on the received video information and count the traffic flow in different sections and time periods; A4. Automatically adjust the phase and duration of traffic lights according to the real-time location and driving direction of vehicles to ensure orderly traffic flow; A5. Announce the congestion and emergency situations of each road section through electronic display screens; A6. Use high-definition cameras to monitor the real-time conditions of each road section and resolve emergencies.

Images