T-shaped intersection left-turn guiding system based on road right dynamic allocation and self-adaptive control

Abstract

The invention discloses a T-shaped intersection left-turn guiding system based on road right dynamic allocation and adaptive control. The system comprises a video monitoring module, a road right allocation module, a guiding control module and a guiding illumination module. The guiding control module receives the left turning signal in real time, calculates a corresponding control mode signal and sends the control mode signal to the guiding illumination module; the guiding illumination module is switched among a plurality of guiding modes according to the control mode signal to realize dynamic visual guiding in the left turning process of the vehicle; the video monitoring module collects an intersection operation video in real time, carries out license plate number identification on the collected video, and sends a vehicle type, a license plate number and a mileage to the right of way distribution module. The road right distribution module calculates the road right quotas of the current vehicles based on the vehicle types, the road right quotas are ranked from high to low, and the vehicles ranked in the front order pass preferentially. According to the invention, integrated coordination of adaptive control of traffic flow, vehicle guidance and safety management and control can be realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of intelligent traffic management, and particularly relates to a left-turn guiding system for T-shaped intersection based on dynamic allocation of road right and adaptive control. BACKGROUND

[0002] With the continuous growth of urban motor vehicle ownership, the uneven occupation of road resources and traffic congestion caused by traffic travel are increasingly prominent. The traditional traffic restriction and congestion management mode cannot dynamically adjust according to real-time traffic load, vehicle type difference and road running state, resulting in low resource utilization and insufficient vehicle passing efficiency.

[0003] In addition, in complex intersection scenarios, especially in the left-turn passing process of T-shaped intersection, vehicles are prone to line crossing, road grabbing and scraping with the central separation belt due to the lack of effective flow guiding and visual cues, significantly increasing the traffic safety risk. The existing left-turn guiding signs are mostly static marking lines or low-brightness display forms, which are difficult to provide effective visual guidance for drivers in low-visibility environments such as rain, fog, etc. At the same time, the image recognition accuracy of traditional cameras in such environments is limited, and there are errors in license plate recognition and vehicle matching, and there is a lack of voice prompts and interactive feedback functions, which cannot meet the needs of intelligent traffic management.

[0004] Therefore, in order to solve the above problems, a left-turn guiding system for T-shaped intersection based on dynamic allocation of road right and adaptive control is needed, which can realize the integrated cooperation of adaptive control of traffic flow, vehicle guiding and safety control. SUMMARY

[0005] Therefore, the purpose of the present application is to overcome the defects in the prior art and provide a left-turn guiding system for T-shaped intersection based on dynamic allocation of road right and adaptive control, which can realize the integrated cooperation of adaptive control of traffic flow, vehicle guiding and safety control.

[0006] The left-turn guiding system for T-shaped intersection based on dynamic allocation of road right and adaptive control of the present application comprises a video monitoring module, a road right allocation module, a guiding control module and a guiding lighting module.

[0007] The guiding control module is used for receiving the left-turn signal in real time, calculating the corresponding control mode signal, and sending the control mode signal to the guiding lighting module. The guiding lighting module switches between several guiding modes according to the control mode signal, realizing the dynamic visual guidance of the vehicle left-turn process.

[0008] The video monitoring module is used for collecting intersection operation video in real time, performing license plate number recognition on the collected video, and sending vehicle type, license plate number and driving mileage to the road right allocation module; the road right allocation module calculates the road right quota of the current vehicle based on the vehicle type, sorts the road right quota from high to low, and the vehicles in the front sequence have priority to pass.

[0009] Further, the guide lighting module comprises a variable light-emitting strip-shaped lamp; wherein the variable light-emitting strip-shaped lamp is installed directly below the outside of the left-turn guide sign line on the road surface.

[0010] Further, the left-turn signal is received in real time, and the corresponding control mode signal is calculated, specifically including:

[0011] S11. Real-time receiving left-turn signal, in a period T, when the left-turn signal is red light and the remaining time of red light is greater than or equal to 3s, mode three signal is sent to the guide lighting module, at this time the guide lighting module executes mode three, that is, the variable light-emitting strip-shaped lamp displays red light, otherwise S12 is executed;

[0012] S12. When the left-turn signal is red light and the remaining time of red light is less than 3s, mode two signal is sent to the guide lighting module, at this time the guide lighting module executes mode two, that is, the variable light-emitting strip-shaped lamp displays yellow light, otherwise S13 is executed;

[0013] S13. When the left-turn signal is yellow light, mode two signal is sent to the guide lighting module, at this time the guide lighting module executes mode two, that is, the variable light-emitting strip-shaped lamp displays yellow light, otherwise S14 is executed;

[0014] S14. When the left-turn signal is green light and the remaining time of green light is greater than or equal to 3s, mode one signal is sent to the guide lighting module, at this time the guide lighting module executes mode one, that is, the variable light-emitting strip-shaped lamp displays green light, otherwise S15 is executed;

[0015] S15. When the left-turn signal is green light and the remaining time of green light is less than 3s, mode two signal is sent to the guide lighting module, at this time the guide lighting module executes mode two, that is, the variable light-emitting strip-shaped lamp displays yellow light, otherwise S11 is executed.

[0016] Further, the variable light-emitting strip-shaped lamp comprises an LED light source, a light control device, a light distribution component and a mounting shell;

[0017] ​​​​The road surface part is a rectangular high-strength transparent glass material, and the installation shell is a cuboid, which is made of engineering plastic, and the installation shell cover is made of PC material, and the installation shell is provided with a circular hole slot around the installation shell cover, and the light control device is responsible for device power supply, and the light distribution part controls the light color and is divided into three operation modes.

[0018] Further, the video monitoring module comprises a camera and an electronic police snapshot integrated body; and the video monitoring module is installed at the top end of the F-pole of the intersection.

[0019] Further, the collected video is subjected to license plate number recognition, specifically comprising:

[0020] S21. Importing a real-time collected video signal and converting it into an image I;

[0021] S22. Dividing the image I into three channel images, and taking the image with the smallest gray scale as I(x);

[0022] After the image I(x) is subjected to defogging optimization, an image g(x) is obtained;

[0023] g(x)=(I(x)-B) / max[H(x),H0]+B, B is an atmospheric light component characteristic value, H(x) is a light transmittance, and H0 is a threshold value, when H(x)<H0, H(x)=H0;

[0024] Then, the g(x) is superimposed with the other two channel images to obtain an optimized image I1;

[0025] S23. Dividing the image I1 into three channel images, namely IR, IG and IB, each channel image is subjected to local histogram equalization processing by calling a GetLocalHisteq function, and the local area histogram is processed in an adaptive manner, and then the enhanced component images IR1, IG2 and IB3 returned by the GetLocalHisteq function are obtained, and the three enhanced component images are recombined into an RGB image I2;

[0026] S24. The image I2 is subjected to Gaussian filtering to obtain images under different scales, and single-scale Retinex calculation is performed on the images under each scale, and a particle swarm optimization algorithm is used to optimize the Gaussian surrounding space scale parameter , and the optimal value is calculated;

[0027] Fitness function of the particle swarm algorithm ; D is the image brightness value of the image I2, the corresponding scale of the reflection component is obtained, and the reflection components under all scales are subjected to weighted average to obtain the final reflection component, and the final reflection component is fused with the image I2 to obtain the defogged image I3;

[0028] S24. Recognize the image I3 using yolov5, output the license plate number data.

[0029] Further, the road right quota of the current vehicle is calculated according to the following formula :

[0030] ;

[0031] Wherein, is the initial value of the road right quota; is the road right quantity consumed by the unit driving distance of the first type of vehicle; is the driving distance of the first type of vehicle.

[0032] Further, it further comprises a display notification module; the display notification module synchronously displays the vehicle road right quota and the current passing state.

[0033] Further, it further comprises a voice prompt module; the voice prompt module provides voice guidance or early warning prompt, realizes real-time interaction and safety auxiliary of the driver.

[0034] The beneficial effects of the present application are: the T-shaped intersection left turn guiding system based on dynamic allocation and adaptive control of road right disclosed by the present application realizes intelligent guidance and dynamic passing management of vehicles in the left turn process through the construction of a multi-module collaborative architecture of video monitoring, road right allocation, guiding control and guiding lighting. The video monitoring module collects the intersection vehicle information in real time, identifies the license plate and the vehicle type; the road right allocation module calculates the road right quota according to the vehicle type and the driving parameter and sorts it to determine the priority passing sequence; the guiding control module generates mode instructions according to the left turn signal, drives the guiding lighting module to adaptively switch between multiple guiding modes, realizes the precise visual guidance and order release of the left turn vehicle, and improves the intersection passing efficiency and safety. BRIEF DESCRIPTION OF DRAWINGS

[0035] The present application will be further described below in conjunction with the drawings and examples:

[0036] Figure 1 It is a schematic diagram of the principle structure of the T-shaped intersection left turn guiding system of the present application. DETAILED DESCRIPTION

[0037] The present application will be further described below in conjunction with the drawings and examples:

[0038] The present embodiment discloses a T-shaped intersection left turn guiding system based on dynamic allocation and adaptive control of road right, which comprises a video monitoring module, a road right allocation module, a guiding control module and a guiding lighting module.

[0039] The guiding control module is configured to receive a left-turn signal in real time, calculate a corresponding control mode signal, and send the control mode signal to the guiding lighting module; the guiding lighting module switches between a plurality of guiding modes according to the control mode signal, so as to realize dynamic visual guiding of the left-turn process of the vehicle.

[0040] The video monitoring module is configured to collect intersection operation videos in real time, perform license plate number recognition on the collected videos, and send the vehicle type, license plate number and driving mileage to the road right allocation module; the road right allocation module calculates the road right quota of the current vehicle based on the vehicle type, sorts the road right quotas from high to low, and the vehicles in the front sequence are given priority to pass.

[0041] In this embodiment, the guiding lighting module includes a variable light-emitting bar lamp; wherein the variable light-emitting bar lamp is installed directly below the outside of the left-turn guiding mark line on the road surface.

[0042] The variable light-emitting bar lamp includes an LED light source, a light control device, a light distribution component and a mounting shell; the road surface part is a rectangular high-strength transparent glass material with a size of 2000mm*150mm, the mounting shell is a cuboid with an outer size of 2000mm*150mm*9mm, the material is engineering plastic, the mounting shell cover is made of PC material, circular hole grooves are left around the mounting shell, the light control device is responsible for device power supply, and the light distribution component controls the light color of the LED light source and is divided into three operation modes. Among them, mode one is green light, mode two is yellow light, and mode three is red light.

[0043] By setting the variable light-emitting bar lamp outside the left-turn guiding mark line on the road surface, dynamic visual presentation of the left-turn guiding information is realized. By using the high-strength transparent glass and engineering plastic integrated packaging structure, good waterproof, pressure resistance and impact resistance are achieved, which is suitable for complex road environment. The LED light source cooperates with the light distribution component to realize multi-mode light control, automatically switches the color and flashing frequency according to the guiding control signal, thereby significantly improving the recognizability and safety of vehicle left-turn guiding in night, rain, fog and low visibility conditions, and enhancing the overall guiding effect and reliability of the system.

[0044] In this embodiment, the guiding control module receives the left-turn signal sent by the signal lamp control module in real time, and calculates the corresponding control mode signal, wherein the signal lamp control module includes a signal lamp controller arranged at the intersection, the signal lamp controller sends the left-turn signal of the signal lamp to the guiding control module in real time, and the left-turn signal includes the color and time field of the left-turn signal lamp. The guiding control module includes CPU and the like.

[0045] The cooperation process between the guiding control module and the guiding lighting module specifically includes:

[0046] S11. The guidance control module receives the left turn signal in real time. In a period T, when the left turn signal is red and the remaining time of the red light is greater than or equal to 3s, a mode three signal is sent to the guidance lighting module. At this time, the guidance lighting module executes mode three, that is, the variable light-emitting bar-shaped lamp displays red light. Otherwise, S12 is executed. The period T can be set according to the actual working condition, and will not be described here.

[0047] S12. When the left turn signal is red and the remaining time of the red light is 0<3s, a mode two signal is sent to the guidance lighting module. At this time, the guidance lighting module executes mode two, that is, the variable light-emitting bar-shaped lamp displays yellow light. Otherwise, S13 is executed.

[0048] S13. When the left turn signal is yellow, a mode two signal is sent to the guidance lighting module. At this time, the guidance lighting module executes mode two, that is, the variable light-emitting bar-shaped lamp displays yellow light. Otherwise, S14 is executed.

[0049] S14. When the left turn signal is green and the remaining time of the green light is greater than or equal to 3s, a mode one signal is sent to the guidance lighting module. At this time, the guidance lighting module executes mode one, that is, the variable light-emitting bar-shaped lamp displays green light. Otherwise, S15 is executed.

[0050] S15. When the left turn signal is green and the remaining time of the green light is 0<3s, a mode two signal is sent to the guidance lighting module. At this time, the guidance lighting module executes mode two, that is, the variable light-emitting bar-shaped lamp displays yellow light. Otherwise, S11 is executed.

[0051] Through the cooperative logic of the guidance control module and the guidance lighting module, multi-mode dynamic light guidance control based on the signal light state and the remaining time is realized. The system can automatically switch the light-emitting color in different stages of red, yellow and green lights, form a visual prompt sequence of red stop, yellow warning and green go, and effectively improve the early perception ability of the driver to the change of the left turn signal. The adaptive light switching mechanism not only optimizes the start and waiting rhythm of the vehicle left turn, but also enhances the safety and fluency of the intersection traffic in complex traffic environment, significantly reduces the risk of overrunning and rushing.

[0052] In the embodiment, the video monitoring module includes a camera and an electronic police snapshot. The video monitoring module is installed at the top of the F-pole of the intersection. The camera is a 360-degree spherical infrared camera, which collects video signals in real time and sends them to the storage module through the data transmission module. The data transmission module is composed of an electronic antenna, a switching interface, an optical fiber cable and a transmitter, and is responsible for data transmission. The storage module is composed of a hard disk, and is responsible for all data storage and recordation.

[0053] ​​​​By installing a video monitoring module with integrated camera and electronic police snapshot function on the top of the F pole at the intersection, high-level monitoring of the intersection panorama and key passing areas is realized. This layout has wide view and few blind spots, and can collect vehicle running state and license plate information in real time, providing accurate data support for road right allocation and traffic violation identification, thereby improving system monitoring accuracy and the reliability of law enforcement evidence.

[0054] In this embodiment, the collected video is subjected to license plate number recognition, specifically including:

[0055] S21. Importing real-time collected video signals and converting them into images I;

[0056] S22. Dividing the image I into 3-channel images, and taking the image with the smallest gray scale as I(x);

[0057] Implementing image dark channel defogging processing, and obtaining image g(x) after defogging optimization of the image I(x);

[0058] g(x) = (I(x) - B) / max[H(x), H0] + B, B is the characteristic value of atmospheric light component, the value of B is the average value of the brightness of the 0.1% pixels with the highest brightness in the image I(x) at the corresponding position of the image I; H(x) is the transmittance and is a constant; H0 is the threshold value, which is 0.1 here; when H(x) < H0, let H(x) = H0;

[0059] Then, g(x) is superimposed with the other two channel images to obtain the optimized image I1;

[0060] S23. Dividing the image I1 into 3-channel images, namely IR, IG and IB, each of which is subjected to local histogram equalization processing by calling the GetLocalHisteq function, and the local region histogram is processed in an adaptive manner, and then the enhanced component images IR1, IG2 and IB3 returned by the GetLocalHisteq function are obtained, and the three enhanced component images are recombined into an RGB image I2;

[0061] By adjusting the histogram distribution of the image, the contrast and details of the image are enhanced;

[0062] S24. Gaussian filtering is performed on the image I2 to obtain images at different scales, and single-scale Retinex calculation is performed on the images at each scale, and the particle swarm optimization algorithm is used to optimize the spatial scale parameter around the Gaussian function, and the optimal value is calculated;

[0063] Fitness function of the particle swarm algorithm ; MSE represents mean square error, used to measure the difference between two images, here representing the error between the filtered image and the original image brightness; D is the image brightness value of image I2, the reflection component under the corresponding scale is obtained, and then the reflection components under all scales are weighted and averaged to obtain the final reflection component, and the final reflection component is fused with the image I2 to obtain the defogged image I3;

[0064] S24. The image I3 is identified by yolov5, and the license plate number data is output. Wherein, the YOLOv5 algorithm is based on a deep convolutional neural network structure, and realizes high-precision positioning of vehicle and license plate region in the image through steps of feature extraction, feature fusion and multi-scale detection. The algorithm uses CSPDarknet backbone network to extract multi-level feature information in the feature extraction stage, uses FPN and PAN structure to enhance the detection ability under different scales in the feature fusion stage, and generates a detection frame containing the license plate position, category and confidence in the detection output stage. Through character recognition processing on the detection frame region, the corresponding license plate number data is finally output, realizing high-precision extraction of license plate information in complex environment.

[0065] In the embodiment, the road right quota of the current vehicle is calculated according to the following formula :

[0066] ;

[0067] Wherein, is the initial value of the road right quota, which can be set according to the actual working condition, such as 1000000; is the road right quantity consumed by the first type vehicle per unit driving mileage. The vehicle type can be divided into five categories: small car (7 seats and below), medium car (8 to 20 seats), large car (21 seats and above), small truck (load 3.5 tons and below), and medium and large truck (load 3.5 tons and above). Taking small car as the benchmark, the corresponding road right quantity is set to 1.0, and other types of vehicles are converted according to the "equivalent small car number" or "road right weight", for example, the medium car corresponds to 1.5 to 2.0, the large car corresponds to 2.5 to 3.0, the small truck corresponds to 1.8 to 2.5, and the medium and large truck corresponds to 3.0 to 4.5; is the driving mileage of the first type vehicle.

[0068] The left turn guiding system of the application further comprises a display notification module; the display notification module is composed of a liquid crystal display and is installed at the top of the left turn exit column at the intersection; the display notification module can synchronously display the vehicle road right quota and the current traffic state.

[0069] The voice prompt module is composed of a high-pitched loudspeaker. When the guiding illumination module implements mode one, the voice prompt is "please pass quickly"; when the guiding illumination module implements mode two, the voice prompt is "please observe passing"; and when the guiding illumination module implements mode three, the voice prompt is "prohibit passing".

[0070] By setting the display notification module and the voice prompt module, dual guidance of vision and hearing is realized. The display notification module can synchronously display the vehicle right-of-way quota and the passing state, improving the real-time perception of the driver; the voice prompt module dynamically broadcasts the passing instruction according to the guiding mode, so that the driver can still accurately obtain traffic information in a limited vision or noise environment, thereby significantly improving the safety and interactivity of left-turn guiding.

[0071] Finally, it should be pointed out that the above embodiments are only used to illustrate the technical solutions of the present application and not to limit the present application. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced equivalently without departing from the purpose and scope of the present application, and they should be covered in the scope of the claims of the present application.

Claims

1. A left-turn guidance system for T-junctions based on dynamic right-of-way allocation and adaptive control, characterized in that: It includes a video monitoring module, a right-of-way allocation module, a guiding control module, and a guiding lighting module; The guiding control module is used to receive the left-turn signal in real time, calculate the corresponding control mode signal, and send the control mode signal to the guiding lighting module; the guiding lighting module switches between several guiding modes according to the control mode signal to achieve dynamic visual guidance during the vehicle left-turn process; The video monitoring module is used to collect the intersection operation video in real time, identify the license plate number of the collected video, and send the vehicle type, license plate number, and driving mileage to the right-of-way allocation module; the right-of-way allocation module calculates the right-of-way quota of the current vehicle based on the vehicle type, sorts the right-of-way quotas from high to low, and the vehicles ranked in the front have priority to pass.

2. The T-junction left-turn guidance system based on dynamic right-of-way allocation and adaptive control according to claim 1, characterized in that: The guiding lighting module includes variable luminous strip lights; among them, the variable luminous strip lights are installed directly below the outside of the left-turn guiding sign line on the road surface.

3. The T-junction left-turn guidance system based on dynamic right-of-way allocation and adaptive control according to claim 2, characterized in that: Receiving the left-turn signal in real time and calculating the corresponding control mode signal specifically includes: S11. Receive left turn signals in real time. Within one cycle T, when the left turn signal is red and the red light has remaining time... When the time is ≥3s, a mode 3 signal is sent to the guide lighting module. At this time, the guide lighting module executes mode 3, which means the light bar will display red light. Otherwise, execute S12. S12. When the left turn signal is red and the remaining red light time is 0 < If less than 3 seconds have elapsed, a mode 2 signal is sent to the guide lighting module. At this time, the guide lighting module executes mode 2, which means the light bar will display yellow light. Otherwise, S13 is executed. S13. When the left-turn signal is a yellow light, send a mode two signal to the guiding lighting module. At this time, the guiding lighting module executes mode two, that is, the variable luminous strip lights display yellow light, otherwise execute S14; S14. When the left turn signal is green and the green light time remains... When the time is ≥3s, a mode 1 signal is sent to the guide lighting module. At this time, the guide lighting module executes mode 1, which means the light bar will display green light. Otherwise, execute S15. S15. When the left turn signal is green and the remaining green light time is 0 < If the signal is less than 3 seconds, a mode 2 signal is sent to the guide lighting module. At this time, the guide lighting module executes mode 2, which means the light bar will display yellow light. Otherwise, S11 is executed.

4. The T-junction left-turn guidance system based on dynamic right-of-way allocation and adaptive control according to claim 2, characterized in that: The variable luminous strip lights include an LED light source, a lighting control device, a light distribution component, and an installation shell; The road surface part is made of rectangular high-strength transparent glass. The installation shell has an outer cuboid shape and is made of engineering plastic. The installation shell mask is made of PC material. There are circular installation holes left around the installation shell. The lighting control device is responsible for device power supply. The light distribution component and the LED light source control the light color and are divided into 3 operating modes.

5. The T-junction left-turn guidance system based on dynamic right-of-way allocation and adaptive control according to claim 1, characterized in that: The video monitoring module includes a camera and an integrated electronic police capture; the video monitoring module is installed at the top of the intersection F pole.

6. The T-junction left-turn guidance system based on dynamic right-of-way allocation and adaptive control according to claim 5, characterized in that: Identifying the license plate number of the collected video specifically includes: S21. Import the real-time collected video signal and convert it into an image I; S22. Cut the image I into 3-channel images, and use the image with the smallest grayscale as I(x); After dehazing and optimizing the image I(x), obtain the image g(x); g(x) = (I(x) - B) / max[H(x), H0] + B, where B is the atmospheric light component eigenvalue, H(x) is the transmittance, H0 is the threshold. When H(x) < H0, let H(x) = H0; Then superimpose g(x) with the other 2-channel images to obtain the optimized image I1; S23. Divide the image I1 into 3-channel images, namely IR, IG, and IB. Each channel image calls the GetLocalHisteq function for local histogram equalization processing, uses an adaptive method to process the histogram of the local area, and then returns the enhanced component images IR1, IG2, and IB3 processed by the GetLocalHisteq function. Recombine the three enhanced component images into an RGB image I2; S24. Apply Gaussian filtering to image I2 to obtain images at different scales. Perform single-scale Retinex calculations on each scale image and use particle swarm optimization to optimize the spatial scale parameters around the Gaussian filter. Perform optimization and calculate Optimal value; Fitness function of particle swarm optimization algorithm D represents the image brightness value of image I2. The reflection component at the corresponding scale is obtained. Then, the reflection components at all scales are weighted and averaged to obtain the final reflection component. The final reflection component is fused with image I2 to obtain the dehazed image I3. S24. Use yolov5 to identify the image I3 and output the license plate number data.

7. The T-junction left-turn guidance system based on dynamic right-of-way allocation and adaptive control according to claim 1, characterized in that: The current vehicle's right-of-way quota is calculated using the following formula. : ； in, This represents the initial value of the right-of-way quota. For the first The number of road rights consumed per unit distance traveled by each type of vehicle; For the first Mileage of each type of vehicle.

8. The T-junction left-turn guidance system based on dynamic right-of-way allocation and adaptive control according to claim 1, characterized in that: It also includes a display notice module; the display notice module synchronously displays the vehicle right-of-way quota and the current passing status.

9. The T-junction left-turn guidance system based on dynamic right-of-way allocation and adaptive control according to claim 1, characterized in that: It also includes a voice prompt module; the voice prompt module provides voice guidance or warning prompts to enable real-time interaction and safety assistance for the driver.

Images