Vehicle passing management method, system and program product in haze weather

Abstract

The invention belongs to the technical field of traffic management, and particularly discloses a vehicle passing management method, system and program product in hazy weather, and the method comprises the steps: collecting a road monitoring image of a target road section in hazy weather, carrying out the corresponding image filtering processing, obtaining a filtered image after the haze noise is suppressed, and carrying out the wheel detection and vehicle judgment. The method comprises the following steps: firstly, detecting the vehicle density according to the vehicle density, then detecting target vehicles in continuous frames of images by adopting a corresponding target tracking algorithm, judging the vehicle running speed based on different image area positions of the target vehicles, finally judging the traffic jam condition of a target road section according to the vehicle density and the vehicle running speed, and adopting a corresponding traffic control scheme. Through corresponding processing of the road section monitoring image and vehicle identification and detection, the vehicle identification and detection accuracy can be improved, the influence of haze weather can be effectively eliminated, the vehicle passing condition of the road section can be accurately judged, and then targeted vehicle traffic control is performed, and the vehicle passing pressure of the road section can be relieved.

Description

Technical Field

[0001] The present invention belongs to the technical field of traffic management, and particularly relates to a vehicle passing management method, system and program product in haze weather. Background Art

[0002] The visibility of road traffic in haze weather is reduced, which will have a great impact on the driving safety of vehicles and the efficiency of road traffic. Therefore, it is crucial to conduct effective vehicle monitoring and traffic management in haze weather. Efficient vehicle monitoring and traffic management means can timely detect traffic participants on the road, so as to make responses in advance, avoid potential dangers, improve the safety of road traffic, and can also improve the efficiency of road traffic, ensure the smoothness of traffic, and play an important role in aspects such as assisted driving and promoting the development of autonomous driving technology. However, the existing vehicle target monitoring means have problems such as low recognition accuracy and easy omission of detection in low visibility conditions, and it is difficult to accurately control the passing conditions of vehicles and realize the passing management of vehicles in haze weather scenarios. Summary of the Invention

[0003] The purpose of the present invention is to provide a vehicle passing management method, system and program product in haze weather to solve the above problems existing in the prior art.

[0004] To achieve the above purpose, the present invention adopts the following technical solutions: In the first aspect, a vehicle passing management method in haze weather is provided, including: Obtain the road monitoring video stream of the target section in haze weather, and continuously extract road monitoring frame images from the road monitoring video stream; Perform image filtering processing on the road monitoring frame images to obtain filtered monitoring frame images with haze noise suppressed; Use a preset target detection model to detect wheels in the filtered monitoring frame images to obtain wheel detection results; Determine the detected vehicles and their image region position information according to the wheel detection results, and determine the number of detected vehicles in the filtered monitoring frame images; Determine vehicle density data according to the number of detected vehicles and the image size of the filtered monitoring frame images; Based on the target tracking algorithm, determine the image region position information of the detected vehicles in two consecutive filtered monitoring frame images, and determine the vehicle driving speed according to the image region position information of the detected vehicles in two consecutive filtered monitoring frame images; Determine the traffic congestion situation of the target section according to the vehicle density data and the vehicle driving speed, and output a corresponding traffic control plan based on the traffic congestion situation of the target section.

[0005] In a possible design, the image filtering process on the road monitoring frame image to obtain a filtered monitoring frame image with haze noise suppressed includes: Preprocess the road monitoring frame image based on the dark channel prior dehazing algorithm, and perform adaptive filtering on the preprocessed road monitoring frame image using an adaptive filter to obtain a filtered monitoring frame image with haze noise suppressed.

[0006] In a possible design, the adaptive filter calculates the local mean parameter and local variance parameter around each pixel within a sliding window of a fixed size using a set mean equation and variance equation, and calculates the mask parameter between the corresponding pixels within the sliding window based on the local mean parameter and local variance parameter using a set mask parameter equation, and performs pixel-by-pixel masking filtering on the sliding window; the mean equation is

[0007] where represents the local mean parameter, represents a sliding window with an N×M pixel range, and the sliding window uses a sliding window with a 7×7 pixel range, and N = M = 7, and represent any two pixel points within the sliding window, represents and the gray difference between the two pixel points; The variance equation is

[0008] where represents the local variance parameter; The mask parameter equation is

[0009] where represents the mask parameter.

[0010] In a possible design, when the target detection model performs wheel detection on the filtered monitoring frame image, it uses the Hough transform method to detect geometric shapes in the filtered monitoring frame image, and when there are two circles in the detected geometric shapes that meet the set prior conditions, it determines that the two circles are the two wheels of the same vehicle.

[0011] In a possible design, the prior conditions are: the two circles are on the same horizontal line, the ratio of the areas of the two circles is within a set ratio range, the radii of the two circles are both within a set radius range, and the distance between the centers of the two circles is within a set distance range.

[0012] In a possible design, determining the vehicle density data according to the detected vehicle quantity and the image size of the filtered monitoring frame image includes: dividing the detected vehicle quantity by the image size of the filtered monitoring frame image to obtain the vehicle density data.

[0013] In a possible design, determining the image region position information of the detected vehicle in two consecutive filtered monitoring frame images based on the target tracking algorithm and determining the vehicle driving speed according to the image region position information of the detected vehicle in two consecutive filtered monitoring frame images includes: Using the DeepSORT target tracking algorithm to detect the same detected vehicle in two consecutive filtered monitoring frame images, and extracting the image region position information of the same detected vehicle in the two filtered monitoring frame images; Calculating the image distance difference between the two image region position information according to the image region position information of the same detected vehicle in the two filtered monitoring frame images; Calculating the vehicle driving speed according to the inter-frame time interval between the two filtered monitoring frame images and the image distance difference.

[0014] In a second aspect, a vehicle passing management system in a haze weather is provided, including an image acquisition unit, an image filtering unit, a wheel detection unit, a target determination unit, a density determination unit, a speed determination unit, and a congestion management unit, wherein: The image acquisition unit is configured to acquire the road monitoring video stream of the target road section in the haze weather, and continuously extract the road monitoring frame images from the road monitoring video stream; The image filtering unit is configured to perform image filtering processing on the road monitoring frame images to obtain the filtered monitoring frame images after suppressing the haze noise; The wheel detection unit is configured to perform wheel detection on the filtered monitoring frame images by using a preset target detection model to obtain the wheel detection result; The target determination unit is configured to determine the detected vehicle and its image region position information according to the wheel detection result, and determine the number of detected vehicles in the filtered monitoring frame images; The density determination unit is configured to determine the vehicle density data according to the detected vehicle quantity and the image size of the filtered monitoring frame images; The speed determination unit is configured to determine the image region position information of the detected vehicle in two consecutive filtered monitoring frame images based on the target tracking algorithm, and determine the vehicle driving speed according to the image region position information of the detected vehicle in two consecutive filtered monitoring frame images; The congestion management unit is configured to determine the traffic congestion condition of the target road section according to the vehicle density data and the vehicle driving speed, and output a corresponding traffic control plan based on the traffic congestion condition of the target road section.

[0015] In a third aspect, a vehicle traffic management system in fog and haze weather is provided, including: A memory for storing instructions; A processor for reading the instructions stored in the memory and executing any of the methods described in the first aspect according to the instructions.

[0016] In a fourth aspect, a computer-readable storage medium is provided. Instructions are stored on the computer-readable storage medium. When the instructions run on a computer, the computer is made to execute any of the methods described in the first aspect. At the same time, a computer program product is also provided. When the computer program product runs on a computer, it executes any of the methods described in the first aspect.

[0017] Beneficial effects: By collecting road monitoring images of a target section in fog and haze weather and performing corresponding image filtering processing, the present invention obtains a filtered image after suppressing fog and haze noise for wheel detection and vehicle determination to determine the vehicle density. Then, a corresponding target tracking algorithm is used to detect target vehicles in consecutive frame images, and the vehicle driving speed is determined based on the different image region positions of the target vehicles. Finally, according to the vehicle density and the vehicle driving speed, the traffic congestion situation of the target section is judged, and corresponding traffic control schemes are adopted, so that vehicle traffic monitoring and intelligent traffic control of the target section can be realized in fog and haze weather. Through the corresponding processing of the section monitoring images and vehicle recognition and detection, the present invention can improve the accuracy of vehicle recognition and detection, effectively eliminate the influence of fog and haze weather, accurately judge the vehicle traffic conditions of the section, and then perform targeted vehicle traffic control to relieve the vehicle traffic pressure of the section. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0019] Figure 1 It is a schematic diagram of the steps of the method in Embodiment 1 of the present invention; Figure 2 It is a schematic diagram of constructing a vehicle template contour in Embodiment 1 of the present invention; Figure 3 It is a schematic diagram of the composition of the system in Embodiment 2 of the present invention Figure 4 It is a schematic diagram of the composition of the system in Embodiment 3 of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0020] It should be noted here that the descriptions of these embodiments are used to help understand the present invention, but do not constitute a limitation on the present invention. The specific structural and functional details disclosed herein are only used to describe the exemplary embodiments of the present invention. However, the present invention can be embodied in many alternative forms and should not be construed as limited to the embodiments set forth herein.

[0021] It should be understood that unless otherwise clearly specified and defined, the term "connection" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and it can be the communication inside two components. For those of ordinary skill in the art, the specific meanings of the above terms in the embodiments can be understood according to specific situations.

[0022] Specific details are provided in the following description to facilitate a complete understanding of the exemplary embodiments. However, those of ordinary skill in the art should understand that the exemplary embodiments can be implemented without these specific details. For example, the device can be shown in a block diagram to avoid making the example unclear with unnecessary details. In other embodiments, well-known processes, structures, and technologies can be shown without unnecessary details to avoid making the embodiments unclear.

[0023] Embodiment 1: This embodiment provides a method for vehicle traffic management in haze weather, as Figure 1 shown, the method includes the following steps: S1. Obtain the road monitoring video stream of the target section in haze weather, and continuously extract road monitoring frame images from the road monitoring video stream.

[0024] Specifically, when implementing, a monitoring camera needs to be installed at the target section. In haze weather, the road monitoring video stream of the target section can be collected through the monitoring camera. After obtaining the road monitoring video stream of the monitoring camera, road monitoring frame images can be continuously extracted from the road monitoring video stream for subsequent image processing.

[0025] S2. Perform image filtering processing on the road monitoring frame images to obtain filtered monitoring frame images with haze noise suppressed.

[0026] Specifically, when implementing, the road monitoring frame images can be preprocessed based on the dark channel prior dehazing algorithm, and an adaptive filter can be used to perform adaptive filtering processing on the preprocessed road monitoring frame images to obtain filtered monitoring frame images with haze noise suppressed.

[0027] The core idea of the dark channel prior dehazing algorithm is to utilize the scattering characteristics of the atmosphere on light. By estimating the intensity of the atmospheric light and the distance relationship between pixel points, the haze information in the image is inferred, and this information is used to remove the haze in the image, thereby significantly improving the visual effect of the image and making it clearer and brighter. The specific content includes: 1. Dark channel prior theory: In the case of haze-free, at least one color channel in the local area of the image has a very low intensity value. That is, the dark channel of a haze-free image can be expressed as the minimum value of the minimum color channel values within the local window around each pixel point.

[0028] 2. Estimation of transmittance: Through the dark channel map, the transmittance map can be estimated. The transmittance represents the degree of light loss when passing through the atmosphere. The lower the transmittance, the clearer the image. The estimation of transmittance usually involves methods such as minimum filtering and soft matting of the dark channel map.

[0029] 3. Estimation of atmospheric light: Atmospheric light refers to the bright background light generated by the scattering of light by particles in the atmosphere. In the dehazing algorithm, the atmospheric light is usually regarded as the brightest pixel value in the scene.

[0030] 4. Image restoration: Finally, using the estimated transmittance and atmospheric light values, the haze-free image is calculated inversely through a physical model. This process involves optimizing the transmittance and atmospheric light values to obtain the best dehazing effect.

[0031] The advantage of the dark channel prior dehazing algorithm is that it can more accurately estimate the actual scene in the image, thus obtaining a more realistic and natural dehazing effect. After dehazing the road monitoring frame image through the dark channel prior dehazing algorithm, corresponding filters can be used to filter the image to effectively eliminate noise and enhance the feature expressiveness of the image.

[0032] In image processing, filters are mainly used to suppress the high-frequency components of the image, that is, to smooth the image, or to enhance the low-frequency components of the image, that is, to enhance or detect the edges in the image. Since the additive fog in the image (a weather phenomenon that affects the visual clarity of the image, which makes a uniform or non-uniform fog effect appear on the image, reducing the contrast and visibility of the image) can be regarded as a high-frequency component, a low-pass filter mask can be used to eliminate it. Adaptive filters are a type of low-pass filter that changes its characteristics according to the gray values under the mask. Adaptive filters use the local statistical characteristics of the gray values under the mask, that is, the mean and variance. These local characteristics are based on the reasonable parameters of the adaptive filter and are closely related to the appearance of the image. The mean gives the average gray level of the local area, while the variance gives the average contrast of the local area.

[0033] The adaptive filter of this method calculates the local mean parameter and local variance parameter around each pixel within a sliding window of a fixed size using the set mean equation and variance equation, and calculates the mask parameter between the corresponding pixels within the sliding window based on the local mean parameter and local variance parameter using the set mask parameter equation, and performs pixel-by-pixel mask filtering on the sliding window; the mean equation is

[0034] wherein, represents the local mean parameter, represents the sliding window with an N×M pixel range, and the sliding window adopts a sliding window with a 7×7 pixel range, and N = M = 7, and represent any two pixel points within the sliding window, represents and the gray difference between the two pixel points; The variance equation is

[0035] wherein, represents the local variance parameter; The mask parameter equation is

[0036] wherein, represents the mask parameter.

[0037] S3. Use the preset object detection model to detect the wheels in the filtered monitoring frame image to obtain the wheel detection result.

[0038] Specifically, when implementing, after obtaining the filtered monitoring frame image with haze noise suppressed, input the filtered monitoring frame image into the preset object detection model for wheel detection. The object detection model uses the Hough transform method to detect geometric shapes in the filtered monitoring frame image, and when there are two circles in the detected geometric shapes that meet the set prior conditions, determine that the two circles are the two wheels of the same vehicle. The prior conditions include: the two circles are on the same horizontal line, the ratio of the areas of the two circles is within the set ratio range (i.e., the sizes of the two circles are similar), the radii of the two circles are both within the set radius range (such as within a 2-pixel distance range), and the distance between the centers of the two circles is within the set distance range (such as the distance between the centers of the two circles is within the range of 40 to 55 pixel distances).

[0039] S4. Determine the detected vehicle and its image area position information according to the wheel detection result, and determine the number of detected vehicles in the filtered monitoring frame image.

[0040] In specific implementation, after determining two wheels of the same vehicle based on corresponding prior conditions, a simple vehicle template contour as shown in Figure 2 can be constructed in the image based on the image position information of the two wheels as the image area of the detected vehicle, and its image area position information is determined. Then, based on all wheel determination results, the number of detected vehicles in the filtered monitoring frame image is determined.

[0041] S5. Determine the vehicle density data according to the number of detected vehicles and the image size of the filtered monitoring frame image.

[0042] In specific implementation, after determining the number of detected vehicles in the filtered monitoring frame image, the vehicle density data can be obtained by dividing the number of detected vehicles by the image size of the filtered monitoring frame image.

[0043] S6. Determine the image area position information of the detected vehicle in two consecutive filtered monitoring frame images based on the target tracking algorithm, and determine the vehicle driving speed according to the image area position information of the detected vehicle in two consecutive filtered monitoring frame images.

[0044] In specific implementation, the DeepSORT target tracking algorithm is used to detect the same detected vehicle in two consecutive filtered monitoring frame images, and the image area position information of the same detected vehicle in the two filtered monitoring frame images is extracted. The DeepSORT target tracking algorithm is a multi-target tracking algorithm based on deep learning. It combines the Kalman filter and the Hungarian algorithm for target state prediction and data association to achieve continuous tracking of the target. The main steps of DeepSORT target tracking include: Object detection: Use a deep learning model (such as YOLO, SSD, etc.) to detect the targets in the frame image, find all the targets in each frame, and generate a bounding box for each target.

[0045] Feature extraction: For each detected target, extract the features in its bounding box, including appearance features (such as color, texture, etc.) and motion features (such as speed, acceleration, etc.). These features will be used for subsequent target matching and tracking.

[0046] Object matching: Use the Kalman filter to predict the position of each target in the next frame, and use the Hungarian algorithm and cascade matching to calculate the matching degree between the targets in the previous and next frames, find the corresponding relationship of the same target in each frame, and assign a unique ID to each tracked target.

[0047] After detecting the same detected vehicle in two consecutive filtered monitoring frame images, the image distance difference between the two image region position information can be calculated according to the image region position information of the same detected vehicle in the two filtered monitoring frame images. Then, based on the inter-frame time interval between the two filtered monitoring frame images and the image distance difference, the driving speed of the vehicle can be calculated.

[0048] S7. Determine the traffic congestion condition of the target road section according to the vehicle density data and the vehicle driving speed, and output the corresponding traffic control plan based on the traffic congestion condition of the target road section.

[0049] Specifically, during implementation, the traffic congestion condition (such as the congestion level) of the target road section can be comprehensively determined according to the factors of the vehicle density data and the vehicle driving speed. Then, the corresponding traffic control plan (including traffic control plans such as traffic restrictions, speed limits, and changing the road traffic direction) can be retrieved according to the traffic congestion condition of the target road section and output to the corresponding traffic management terminal to achieve targeted traffic control of the target road section, improve the congestion condition of the target road section, and reduce the traffic pressure under hazy weather.

[0050] Embodiment 2: This embodiment provides a vehicle passing management system under hazy weather, as Figure 3 shown, including an image acquisition unit, an image filtering unit, a wheel detection unit, a target determination unit, a density determination unit, a speed determination unit, and a congestion management unit, where: The image acquisition unit is used to acquire the road monitoring video stream of the target road section under hazy weather and continuously extract the road monitoring frame images from the road monitoring video stream; The image filtering unit is used to perform image filtering processing on the road monitoring frame images to obtain the filtered monitoring frame images after suppressing the haze noise; The wheel detection unit is used to perform wheel detection on the filtered monitoring frame images by using a preset target detection model to obtain the wheel detection results; The target determination unit is used to determine the detected vehicle and its image region position information according to the wheel detection results, and determine the number of detected vehicles in the filtered monitoring frame images; The density determination unit is used to determine the vehicle density data according to the number of detected vehicles and the image size of the filtered monitoring frame images; The speed determination unit is used to determine the image region position information of the detected vehicle in two consecutive filtered monitoring frame images based on the target tracking algorithm, and determine the vehicle driving speed according to the image region position information of the detected vehicle in two consecutive filtered monitoring frame images; The congestion management unit is used to determine the traffic congestion condition of the target road section according to the vehicle density data and the vehicle driving speed, and output the corresponding traffic control plan based on the traffic congestion condition of the target road section.

[0051] Embodiment 3: This embodiment provides a vehicle passing management system under smoggy weather, as Figure 4 shown. At the hardware level, it includes: A data interface for establishing data docking between the processor, monitoring cameras, and traffic management terminals; A memory for storing instructions; A processor for reading the instructions stored in the memory and executing the vehicle passing management method under smoggy weather in Embodiment 1 according to the instructions.

[0052] Optionally, the system further includes an internal bus. The processor, memory, and data interface can be interconnected through the internal bus. The internal bus can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, or an EISA (Extended Industry Standard Architecture) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc.

[0053] The memory can but is not limited to including a random access memory (RAM), a read-only memory (ROM), a flash memory, a first input first output (FIFO) memory, and / or a first in last out (FILO) memory, etc. The processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0054] Embodiment 4: This embodiment provides a computer-readable storage medium, on which instructions are stored. When the instructions run on a computer, the computer is caused to execute the vehicle passing management method in Embodiment 1. Among them, the computer-readable storage medium refers to a carrier for storing data, which may include, but is not limited to, floppy disks, optical discs, hard disks, flash memories, USB flash drives, and / or memory sticks, etc. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.

[0055] This embodiment also provides a computer program product. When the computer program product runs on a computer, it executes the vehicle passing management method in Embodiment 1. Among them, the computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.

[0056] Finally, it should be noted that the above are only the preferred embodiments of the present invention and are not intended to limit the protection scope of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A vehicle traffic management method in haze weather, characterized in that: include: Obtain a road monitoring video stream of a target road section under haze weather, and continuously extract road monitoring frame images from the road monitoring video stream; Performing image filtering processing on the road monitoring frame image to obtain a filtered monitoring frame image after suppressing haze noise; Using a preset target detection model to perform wheel detection on the filtered monitoring frame image, obtaining a wheel detection result; Determine the detected vehicle and its image area position information according to the wheel detection result, and determine the number of detected vehicles in the filtered monitoring frame image; Determine vehicle density data based on the number of detected vehicles and the image size of the filtered monitoring frame image; Determine the image area position information of the detected vehicle in two consecutive filtered monitoring frame images based on the target tracking algorithm, and determine the vehicle speed according to the image area position information of the detected vehicle in two consecutive filtered monitoring frame images; The traffic congestion situation of the target road section is determined according to the vehicle density data and the vehicle driving speed, and the corresponding traffic control plan is output based on the traffic congestion situation of the target road section.

2. A vehicle traffic management method in haze weather according to claim 1, characterized in that: The performing of image filtering on the road monitoring frame image to obtain a filtered monitoring frame image after suppressing haze noise includes: The road monitoring frame images are preprocessed based on the dark channel prior defogging algorithm, and an adaptive filter is used to perform adaptive filtering on the preprocessed road monitoring frame images to obtain filtered monitoring frame images with haze noise suppressed.

3. The vehicle traffic management method in haze weather according to claim 2 is characterized in that: The adaptive filter uses the set mean equation and variance equation to calculate the local mean parameter and local variance parameter around each pixel in the sliding window of fixed size, and uses the set mask parameter equation based on the local mean parameter and local variance parameter to calculate the mask parameter between the corresponding pixels in the sliding window, and performs pixel-by-pixel mask filtering on the sliding window based on the mask parameter; the mean equation is in, Characterize the local mean parameter, The sliding window represents the N×M pixel range. The sliding window uses a 7×7 pixel range sliding window, and N=M=7. and Represents any two pixels in the sliding window, Characterization and The grayscale difference between two pixels; The variance equation is in, Characterize the local variance parameter; The mask parameter equation is: in, Characterize mask parameters.

4. The vehicle traffic management method in haze weather according to claim 1 is characterized in that: When the target detection model performs wheel detection on the filtered monitoring frame image, the Hough transform method is used to detect geometric figures in the filtered monitoring frame image, and when there are two circles in the detected geometric figures that meet the set prior conditions, the two circles are determined to be two wheels of the same vehicle.

5. The vehicle traffic management method in haze weather according to claim 4 is characterized in that: The prior conditions are: the two circles are on the same horizontal line, the area ratio of the two circles is within a set ratio range, the radii of the two circles are within a set radius range, and the distance between the centers of the two circles is within a set distance range.

6. The vehicle traffic management method in haze weather according to claim 1 is characterized in that: Determining the vehicle density data according to the number of detected vehicles and the image size of the filtered monitoring frame image includes: dividing the number of detected vehicles by the image size of the filtered monitoring frame image to obtain the vehicle density data.

7. The vehicle traffic management method in haze weather according to claim 1 is characterized in that: The method of determining the image area position information of the detected vehicle in two consecutive filtered monitoring frame images based on the target tracking algorithm, and determining the vehicle speed according to the image area position information of the detected vehicle in two consecutive filtered monitoring frame images, includes: The DeepSORT target tracking algorithm is used to detect the same detected vehicle in two consecutive filtered monitoring frame images, and the image area position information of the same detected vehicle in the two filtered monitoring frame images is extracted; Calculate the image distance difference between the two image area position information according to the image area position information of the same detected vehicle in the two filtered monitoring frame images; The vehicle speed is calculated based on the inter-frame time interval and image distance difference between two filtered monitoring frame images.

8. A vehicle traffic management system in haze weather, characterized in that: It includes an image acquisition unit, an image filtering unit, a wheel detection unit, a target determination unit, a density determination unit, a speed determination unit and a congestion management unit, wherein: An image acquisition unit, used to acquire a road monitoring video stream of a target road section in haze weather, and continuously extract road monitoring frame images from the road monitoring video stream; An image filtering unit is used to perform image filtering processing on the road monitoring frame image to obtain a filtered monitoring frame image after suppressing haze noise; The wheel detection unit is used to perform wheel detection on the filtered monitoring frame image using a preset target detection model to obtain a wheel detection result; The target determination unit is used to determine the detected vehicle and its image area position information according to the wheel detection result, and determine the number of detected vehicles in the filtered monitoring frame image; A density determination unit, used to determine vehicle density data according to the number of detected vehicles and the image size of the filtered monitoring frame image; A speed determination unit, used to determine the image area position information of the detected vehicle in two consecutive filtered monitoring frame images based on a target tracking algorithm, and determine the vehicle's driving speed according to the image area position information of the detected vehicle in two consecutive filtered monitoring frame images; The congestion management unit is used to determine the traffic congestion situation of the target road section according to the vehicle density data and the vehicle driving speed, and output a corresponding traffic control plan based on the traffic congestion situation of the target road section.

9. A vehicle traffic management system in haze weather, characterized in that: include: A memory for storing instructions; A processor is used to read the instructions stored in the memory and execute the vehicle traffic management method in haze weather according to any one of claims 1 to 7 according to the instructions.

10. A computer program product, characterized in that When the computer program product is run on a computer, the vehicle traffic management method in haze weather described in any one of claims 1 to 7 is executed.

Images