Mine laneway inspection target tracking method and system based on video identification

Abstract

The invention relates to the field of image processing, in particular to a mine laneway inspection target tracking method and system based on video recognition, and the method comprises the steps: collecting a mine laneway inspection video, carrying out the frame extraction and graying processing, and sequentially calculating the local brightness index, noise degree and illumination uniformity of each pixel point, further adaptively determining the dark channel image window radius of each pixel point, and finally realizing accurate defogging enhancement of the inspection image. According to the method, the problem of poor defogging effect caused by a fixed window size in a non-uniform illumination area in a traditional dark channel algorithm is effectively solved, and the image quality and the accuracy and robustness of target tracking in a complex mine environment are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of image processing, and specifically to a method and system for tracking targets during mine roadway inspections based on video recognition. Background Technology

[0002] With the advancement of intelligent coal mine construction, target tracking technology based on video recognition for roadway inspections has become a core means to ensure safe production in mines. Mine roadway environments are characterized by complex features such as low illumination, high dust levels, dynamic uneven lighting, and fog. Traditional manual inspections suffer from low efficiency and numerous blind spots. Therefore, there is an urgent need to develop a defogging enhancement method that integrates an adaptive lighting mechanism. By dynamically adjusting the transmittance estimation window, optimizing the atmospheric light calculation model, and combining edge enhancement and color correction technologies, robust target tracking in complex roadway environments can be achieved, providing intelligent support for safe coal mine production.

[0003] The existing technology addresses the issue that during image enhancement of mine roadway inspection videos, dust in the scene often affects image quality, leading to inaccurate target recognition during target tracking. Furthermore, when using conventional dark channel algorithms to enhance mine roadway inspection video images, the varying brightness across different areas of the scene means that calculating the dark channel image using a fixed-size window is significantly affected by uneven lighting in areas with inconsistent illumination, resulting in inaccurate dark channel calculations. Summary of the Invention

[0004] This invention provides a target tracking method and system for mine roadway inspection based on video recognition, to solve the existing problem: In the process of enhancing video images of mine roadway inspection using existing dark channel algorithms, due to the different brightness of different areas in the scene image, the calculation of dark channel images based on a fixed-size window will be greatly affected by the illumination in areas with uneven lighting, resulting in inaccurate dark channel calculation.

[0005] The present invention provides a target tracking method and system for mine roadway inspection based on video recognition, which adopts the following technical solution: In a first aspect, the present invention provides a method for target tracking in mine roadway inspection based on video recognition, the method comprising the following steps: Collect mine roadway inspection videos, perform frame extraction on the videos, obtain any grayscale inspection image, and set an initial matrix window corresponding to each pixel in the inspection image as the center. Based on the grayscale values ​​of each pixel in the grayscale inspection image and other pixels in its initial matrix window, obtain the local brightness index of each pixel. Based on the gradient value of the grayscale value of each pixel in the grayscale inspection image and the grayscale value of other pixels in its initial matrix window, the noise level of each pixel is obtained. Based on the noise level of each pixel in the grayscale inspection image, combined with the gradient value of the grayscale value of each pixel and other pixels in the initial matrix window and the local brightness index, the illumination uniformity of each pixel is obtained. The initial dark channel image window radius of each pixel is preset, and the adaptive dark channel image window radius of each pixel is obtained by combining the illumination uniformity of each pixel. The dark channel image is calculated based on the adaptive dark channel image window radius of each pixel to complete the dehazing and enhancement of the inspection image. The enhanced inspection image is then used to complete the target tracking of the mine roadway inspection.

[0006] Furthermore, the specific methods for acquiring mine roadway inspection videos, performing frame extraction on the videos to obtain any grayscale inspection image, and setting an initial matrix window include: Collect mine roadway inspection videos, extract frames from the videos to obtain frame images, convert any image to grayscale to obtain the grayscale inspection image, and set an initial matrix window corresponding to each pixel in the inspection image, with a window size of n*n pixels.

[0007] Furthermore, the specific method for obtaining the local brightness index of each pixel based on the grayscale values ​​of each pixel in the grayscale inspected image and other pixels in its initial matrix window includes: ; In the formula, Indicating the first image in the inspection image Local brightness index of each pixel Indicates the first The grayscale value of each pixel Indicates the first The grayscale value of the j-th pixel (excluding the i-th pixel) within the initial matrix window containing the pixel. This represents the number of pixels in the initial matrix window minus 1.

[0008] Furthermore, the specific method for obtaining the noise level of each pixel based on the gradient value of the grayscale value of each pixel in the grayscale inspection image and other pixels in its initial matrix window includes: ; In the formula, Indicates the first The noise level of each pixel Indicates the first The gradient value of the grayscale value of the j-th pixel (excluding the i-th pixel) within the initial matrix window containing all pixels. Indicates the first The average gradient values ​​of the grayscale values ​​of all pixels in the initial matrix window except for the i-th pixel. Indicates the first The initial matrix window containing the nth pixel, excluding the nth pixel... The standard deviation of the gradient values ​​of the grayscale values ​​of pixels other than the one pixel. This represents the number of pixels in the initial matrix window minus 1. Indicates: When When the value is greater than 1, The value is ,when When the value is less than or equal to 1, The value is 1; The gradient value of the grayscale value of the j-th pixel (excluding the i-th pixel) in the initial matrix window containing the i-th pixel is calculated as follows: the absolute value of the difference between the grayscale value of the j-th pixel (excluding the i-th pixel) and the grayscale value of the i-th pixel.

[0009] Furthermore, the method for obtaining the illumination uniformity of each pixel based on the noise level of each pixel in the grayscale inspection image, combined with the gradient value of the grayscale value of each pixel and other pixels in its initial matrix window, and the local brightness index, includes the following specific methods: ; In the formula, Indicates the first Light uniformity of each pixel Indicates the first The noise level of each pixel This represents the number of pixels in the initial matrix window minus 1. Indicates the first The local brightness index of the j-th pixel (excluding the i-th pixel) within the initial matrix window containing the pixels. Indicates the first The gradient value of the grayscale value of the j-th pixel (excluding the i-th pixel) within the initial matrix window containing all pixels. Indicates the first The average local brightness index of all pixels in the initial matrix window except for the i-th pixel.

[0010] Furthermore, the method for obtaining the adaptive dark channel image window radius of each pixel by presetting the initial dark channel image window radius for each pixel and combining it with the illumination uniformity of each pixel includes the following specific methods: The initial dark channel radius of each pixel is preset to be r pixels. The adaptive dark channel image window radius of each pixel is obtained based on the illumination uniformity of each pixel. The specific method is as follows: ; Indicates the first Adaptive dark channel image window radius size per pixel Indicates the first Lighting uniformity of each pixel.

[0011] Furthermore, the specific method for calculating the dark channel image based on the adaptive dark channel image window radius of each pixel to complete the dehazing and enhancement of the inspection image, and using the enhanced inspection image to complete the target tracking of the mine roadway inspection, includes: Based on the adaptive dark channel image window radius of each pixel, pixels in the inspection image whose Euclidean distance to each pixel is less than its adaptive dark channel image window radius are recorded as pixels in the adaptive dark channel image window corresponding to each pixel. The dark channel image is obtained based on the grayscale value of the pixels in the adaptive dark channel image window corresponding to each pixel, thus completing the dark channel dehazing of the inspection image and obtaining the enhanced inspection image. The enhanced inspection image is then used to complete the target tracking of the mine roadway inspection.

[0012] A second aspect of the present invention provides a target tracking system for mine roadway inspection based on video recognition. The system includes a video preprocessing module, a brightness calculation module, a noise calculation module, an illumination uniformity calculation module, a dark passage window size calculation module, and a mine roadway inspection target tracking module, wherein: The video preprocessing module is used to acquire mine roadway inspection videos, perform frame extraction on the videos, obtain any grayscale inspection image, and set an initial matrix window corresponding to each pixel in the inspection image as the center. The brightness calculation module is used to obtain the local brightness index of each pixel based on the grayscale value of each pixel in the grayscale inspection image and other pixels in its initial matrix window. The noise calculation module is used to obtain the noise level of each pixel based on the gradient value of the grayscale value of each pixel in the grayscale inspection image and other pixels in its initial matrix window. The illumination uniformity calculation module is used to obtain the illumination uniformity of each pixel based on the noise level of each pixel in the grayscale inspection image, combined with the gradient value of the grayscale value of each pixel and other pixels in its initial matrix window and the local brightness index. The dark channel window size calculation module is used to preset the initial dark channel image window radius for each pixel, and combine the illumination uniformity of each pixel to obtain the adaptive dark channel image window radius size for each pixel; The mine roadway inspection target tracking module is used to calculate the dark channel image based on the adaptive dark channel image window radius of each pixel, complete the dehazing and enhancement of the inspection image, and use the enhanced inspection image to complete the mine roadway inspection target tracking.

[0013] In a third aspect, the present invention provides a computer-readable storage medium storing a computer program, characterized in that, when the computer program is executed by a processor, it implements the steps of the above-described video recognition-based target tracking method for mine roadway inspection.

[0014] A fourth aspect of the present invention provides a computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor, when executing the computer program, implements the steps of the aforementioned video recognition-based target tracking method for mine roadway inspection.

[0015] The beneficial effects of the technical solution of the present invention are: By analyzing the local brightness, noise level, and illumination uniformity of each pixel, the window size of the dark channel dehazing algorithm is dynamically adjusted, which effectively overcomes the calculation deviation of the traditional fixed window in the uneven illumination area of ​​the mine roadway, and significantly improves the accuracy and robustness of image dehazing. In complex roadway environments with low illumination and high dust, precise illumination-adaptive defogging enhancement effectively restores image details and improves image clarity, providing high-quality image input for subsequent target detection and tracking tasks, thereby improving the accuracy of target identification and tracking during inspection. When calculating illumination uniformity, a noise level factor is introduced to effectively suppress the interference of image noise on illumination evaluation, ensuring stable and reliable operation in real industrial scenarios. By using local window computation and adaptive mechanisms, computational complexity can be controlled while ensuring performance, making it suitable for deployment in embedded devices or edge computing environments and showing good prospects for engineering applications. To address the problem that conventional dark channel algorithms are easily affected by illumination in images with uneven lighting, leading to inaccurate dark channel image calculations, this invention calculates the local brightness index of each pixel, then obtains the illumination uniformity based on the other local brightness indices around the pixel, and then adaptively calculates the window adjustment coefficient of each pixel to obtain an adaptive window size. This achieves the beneficial effect of obtaining more accurate dark channel images for mine roadway images with uneven illumination. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a flowchart illustrating the steps of a target tracking method for mine roadway inspection based on video recognition according to the present invention. Figure 2 This is a structural block diagram of a mine roadway inspection target tracking system based on video recognition according to the present invention. Detailed Implementation

[0018] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a video recognition-based target tracking method and system for mine roadway inspection proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0019] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0020] The following description, in conjunction with the accompanying drawings, details the specific scheme of the target tracking method and system for mine roadway inspection based on video recognition provided by this invention.

[0021] Please see Figure 1 It illustrates the first objective of the present invention, a flowchart of a target tracking method for mine roadway inspection based on video recognition, the method comprising the following steps: Step S001: Collect mine roadway inspection video, perform frame extraction on the video, obtain any grayscale inspection image, and set an initial matrix window corresponding to each pixel in the inspection image as the center.

[0022] Specifically, the process involves acquiring mine roadway inspection videos, performing frame extraction on the videos to obtain any grayscale inspection image, and setting an initial matrix window corresponding to each pixel in the inspection image as the center. The specific method is as follows: Collect mine roadway inspection videos, extract frames from the videos to obtain frame images, convert any image to grayscale to obtain the grayscale inspection image, and set an initial matrix window corresponding to each pixel in the inspection image, with a window size of n*n pixels.

[0023] It should be noted that the present invention does not limit the initial matrix window size. In this embodiment, the initial matrix window size is the minimum window size, where n=3. The initial matrix window size has little impact on the subsequent calculation results.

[0024] Step S002: Based on the grayscale values ​​of each pixel in the grayscale inspection image and other pixels in its initial matrix window, obtain the local brightness index of each pixel.

[0025] It should be noted that for inspection images, due to the limited lighting conditions in mine roadways, different areas are affected by lighting to varying degrees. Therefore, when calculating dark channels for different areas, the degree of lighting influence should be considered. For areas with uneven lighting, a smaller window should be used when calculating dark channel images to avoid the influence of different lighting on the dark channel calculation; while for areas with uniform lighting, a larger window can be used to improve statistical stability.

[0026] It should be further noted that in the inspection image, the brightness of a pixel is mainly related to the gray value of the local pixel. In addition, the brightness also affects the local image contrast. Therefore, the local brightness of the pixel is initially measured based on the gray value of other pixels in the initial matrix window where each pixel is located and the gradient value of the gray value.

[0027] Specifically, based on the grayscale values ​​of each pixel in the grayscale inspected image and other pixels in its initial matrix window, the local brightness index of each pixel is obtained. The specific method is as follows: ; In the formula, Indicating the first image in the inspection image Local brightness index of each pixel Indicates the first The grayscale value of each pixel Indicates the first The grayscale value of the j-th pixel (excluding the i-th pixel) within the initial matrix window containing the pixel. This represents the number of pixels in the initial matrix window minus 1.

[0028] Step S003: Based on the gradient value of the grayscale value of each pixel in the grayscale inspection image and the grayscale value of other pixels in the initial matrix window, obtain the noise level of each pixel.

[0029] It should be noted that, for inspection images, in addition to the local brightness index, the noise that may exist in the image can also have a great impact on the subsequent calculation of illumination uniformity, causing the calculated illumination uniformity to be too high. Therefore, it is necessary to reduce the impact of noise on illumination uniformity.

[0030] Specifically, based on the gradient value of the grayscale value of each pixel in the grayscale inspected image and its grayscale value relative to other pixels in the initial matrix window, the noise level of each pixel is obtained. The specific method is as follows: ; In the formula, Indicates the first The noise level of each pixel Indicates the first The gradient value of the grayscale value of the j-th pixel (excluding the i-th pixel) within the initial matrix window containing all pixels. Indicates the first The average gradient values ​​of the grayscale values ​​of all pixels in the initial matrix window except for the i-th pixel. Indicates the first The initial matrix window containing the nth pixel, excluding the nth pixel... The standard deviation of the gradient values ​​of the grayscale values ​​of pixels other than the one pixel. This represents the number of pixels in the initial matrix window minus 1. Indicates: When When the value is greater than 1, The value is ,when When the value is less than or equal to 1, The value is 1; The gradient value of the grayscale value of the j-th pixel (excluding the i-th pixel) in the initial matrix window containing the i-th pixel is calculated as follows: the absolute value of the difference between the grayscale value of the j-th pixel (excluding the i-th pixel) and the grayscale value of the i-th pixel.

[0031] It should be noted that in the formula This represents the ratio of the difference between the gradient of each pixel's grayscale value and its mean gradient to its standard deviation within an n*n range. If this ratio is greater than one, the pixel's gradient value is considered to be highly outlier and may be a noise point. Based on this, the nth... Calculate the noise level of all other pixels within an n*n range of a given pixel.

[0032] Step S004: Based on the noise level of each pixel in the grayscale inspection image, combined with the gradient value of the grayscale value of each pixel and other pixels in the initial matrix window and the local brightness index, obtain the illumination uniformity of each pixel.

[0033] It should be noted that for any given pixel, it is necessary to determine whether there is uneven local lighting by observing the changes in the local brightness index of other pixels around it.

[0034] Specifically, based on the noise level of each pixel in the grayscale inspected image, combined with the gradient value of the grayscale value of each pixel with other pixels in its initial matrix window and the local brightness index, the illumination uniformity of each pixel is obtained. The specific method is as follows: ; In the formula, Indicates the first Light uniformity of each pixel Indicates the first The noise level of each pixel This represents the number of pixels in the initial matrix window minus 1. Indicates the first The local brightness index of the j-th pixel (excluding the i-th pixel) within the initial matrix window containing the pixels. Indicates the first The gradient value of the grayscale value of the j-th pixel (excluding the i-th pixel) within the initial matrix window containing all pixels. Indicates the first The average local brightness index of all pixels in the initial matrix window except for the i-th pixel.

[0035] It should be noted that in the formula The first part represents gradient reduction. For regions in an image that already have gradients, the larger the gradient, the higher the illumination uniformity. Therefore, this part reduces the impact of the gradient on illumination uniformity. The formula is derived from the first... The illumination uniformity is calculated by analyzing the local brightness index of other pixels around a pixel. If the local brightness index around a pixel differs significantly from the average local brightness index of that area, it indicates that the illumination is uneven and the illumination uniformity is low.

[0036] Step S005: Preset the initial dark channel image window radius for each pixel, and combine it with the illumination uniformity of each pixel to obtain the adaptive dark channel image window radius size for each pixel.

[0037] It should be noted that for any given pixel, the lower its illumination uniformity, the smaller the window used for dark channel calculations to reduce the impact of uneven illumination on the dark channel algorithm, so that the illumination of pixels within the window is as uniform as possible.

[0038] Specifically, the initial dark channel image window radius for each pixel is preset, and the adaptive dark channel image window radius for each pixel is obtained by combining the illumination uniformity of each pixel. The specific method is as follows: ； Indicates the first Adaptive dark channel image window radius size per pixel Indicates the first Lighting uniformity of each pixel.

[0039] It should be noted that the present invention does not specifically limit r. In the present invention, r is taken as an empirical value of 1.5. The specific situation in other embodiments depends on the implementation.

[0040] Step S006: Calculate the dark channel image based on the adaptive dark channel image window radius of each pixel, complete the dehazing and enhancement of the inspection image, and use the enhanced inspection image to complete the target tracking of the mine roadway inspection.

[0041] Specifically, the dark channel image is calculated based on the adaptive dark channel image window radius for each pixel to complete the dehazing and enhancement of the inspection image. The enhanced inspection image is then used to complete the target tracking for mine roadway inspection. The specific method is as follows: Based on the adaptive dark channel image window radius of each pixel, pixels in the inspection image whose Euclidean distance to each pixel is less than its adaptive dark channel image window radius are recorded as pixels in the adaptive dark channel image window corresponding to each pixel. The dark channel image is obtained based on the grayscale value of the pixels in the adaptive dark channel image window corresponding to each pixel, thus completing the dark channel dehazing of the inspection image and obtaining the enhanced inspection image. The enhanced inspection image is then used to complete the target tracking of the mine roadway inspection.

[0042] It should be noted that this invention does not limit the target tracking algorithm for mine roadway inspection. In this embodiment, a neural network is used for calculation. In other embodiments, the target tracking algorithm for mine roadway inspection depends on the specific implementation.

[0043] Please see Figure 2 The diagram illustrates the second objective of this invention: a structural block diagram of a target tracking system for mine roadway inspection based on video recognition. This system includes the following modules: The video preprocessing module is used to acquire mine roadway inspection videos, perform frame extraction on the videos, obtain any grayscale inspection image, and set an initial matrix window corresponding to each pixel in the inspection image as the center. The brightness calculation module is used to obtain the local brightness index of each pixel based on the grayscale value of each pixel in the grayscale inspection image and other pixels in its initial matrix window. The noise calculation module is used to obtain the noise level of each pixel based on the gradient value of the grayscale value of each pixel in the grayscale inspection image and other pixels in its initial matrix window. The illumination uniformity calculation module is used to obtain the illumination uniformity of each pixel based on the noise level of each pixel in the grayscale inspection image, combined with the gradient value of the grayscale value of each pixel and other pixels in its initial matrix window and the local brightness index. The dark channel window size calculation module is used to preset the initial dark channel image window radius for each pixel, and combine the illumination uniformity of each pixel to obtain the adaptive dark channel image window radius size for each pixel; The mine roadway inspection target tracking module is used to calculate the dark channel image based on the adaptive dark channel image window radius of each pixel, complete the dehazing and enhancement of the inspection image, and use the enhanced inspection image to complete the mine roadway inspection target tracking.

[0044] A third objective of this invention is to provide a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the aforementioned video recognition-based target tracking method for mine roadway inspection.

[0045] The fourth objective of this invention is to provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the aforementioned video recognition-based target tracking method for mine roadway inspection.

[0046] This embodiment analyzes the local brightness, noise level, and illumination uniformity of each pixel, and dynamically adjusts the window size of the dark channel dehazing algorithm. This effectively overcomes the calculation deviation of the traditional fixed window in areas with uneven illumination in mine roadways, and significantly improves the accuracy and robustness of image dehazing. In complex roadway environments with low illumination and high dust, precise illumination-adaptive defogging enhancement effectively restores image details and improves image clarity, providing high-quality image input for subsequent target detection and tracking tasks, thereby improving the accuracy of target identification and tracking during inspection. When calculating illumination uniformity, a noise level factor is introduced to effectively suppress the interference of image noise on illumination evaluation, ensuring stable and reliable operation in real industrial scenarios. By using local window computation and adaptive mechanisms, computational complexity can be controlled while ensuring performance, making it suitable for deployment in embedded devices or edge computing environments and showing good prospects for engineering applications. To address the problem that conventional dark channel algorithms are easily affected by illumination in images with uneven lighting, leading to inaccurate dark channel image calculations, this invention calculates the local brightness index of each pixel, then obtains the illumination uniformity based on the other local brightness indices around the pixel, and then adaptively calculates the window adjustment coefficient of each pixel to obtain an adaptive window size. This achieves the beneficial effect of obtaining more accurate dark channel images for mine roadway images with uneven illumination.

[0047] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0048] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0049] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0050] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0051] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for tracking targets during mine roadway inspections based on video recognition, characterized in that, The method includes the following steps: Collect mine roadway inspection videos, perform frame extraction on the videos, obtain any grayscale inspection image, and set an initial matrix window corresponding to each pixel in the inspection image as the center. Based on the grayscale values ​​of each pixel in the grayscale inspection image and other pixels in its initial matrix window, the local brightness index of each pixel is obtained. Based on the gradient value of the grayscale value of each pixel in the grayscale inspection image and the grayscale value of other pixels in its initial matrix window, the noise level of each pixel is obtained. Based on the noise level of each pixel in the grayscale inspection image, combined with the gradient value of the grayscale value of each pixel and other pixels in the initial matrix window and the local brightness index, the illumination uniformity of each pixel is obtained. The initial dark channel image window radius of each pixel is preset, and the adaptive dark channel image window radius of each pixel is obtained by combining the illumination uniformity of each pixel. The dark channel image is calculated based on the adaptive dark channel image window radius of each pixel to complete the dehazing and enhancement of the inspection image. The enhanced inspection image is then used to complete the target tracking of the mine roadway inspection.

2. The method for tracking targets in mine roadway inspection based on video recognition according to claim 1, characterized in that, The specific methods for collecting mine roadway inspection videos, performing frame extraction on the videos to obtain any grayscale inspection image, and setting an initial matrix window include: Collect mine roadway inspection videos, extract frames from the videos to obtain frame images, convert any image to grayscale to obtain the grayscale inspection image, and set an initial matrix window corresponding to each pixel in the inspection image, with a window size of n*n pixels.

3. The method for tracking targets in mine roadway inspection based on video recognition according to claim 1, characterized in that, The specific method for obtaining the local brightness index of each pixel based on the grayscale values ​​of each pixel in the grayscale inspection image and other pixels in its initial matrix window includes: ； In the formula, Indicating the first image in the inspection image Local brightness index of each pixel Indicates the first The grayscale value of each pixel Indicates the first The grayscale value of the j-th pixel (excluding the i-th pixel) within the initial matrix window containing the pixel. This represents the number of pixels in the initial matrix window minus 1.

4. The method for tracking targets in mine roadway inspection based on video recognition according to claim 1, characterized in that, The method for obtaining the noise level of each pixel based on the gradient value of the grayscale value of each pixel in the grayscale inspection image and other pixels in its initial matrix window includes the following specific methods: ； In the formula, Indicates the first The noise level of each pixel Indicates the first The gradient value of the grayscale value of the j-th pixel (excluding the i-th pixel) within the initial matrix window containing all pixels. Indicates the first The average gradient values ​​of the grayscale values ​​of all pixels in the initial matrix window except for the i-th pixel. Indicates the first The initial matrix window containing the nth pixel, excluding the nth pixel... The standard deviation of the gradient values ​​of the grayscale values ​​of pixels other than the one pixel. This represents the number of pixels in the initial matrix window minus 1. Indicates: When When the value is greater than 1, The value is ,when When the value is less than or equal to 1, The value is 1; The gradient value of the grayscale value of the j-th pixel (excluding the i-th pixel) in the initial matrix window containing the i-th pixel is calculated as follows: the absolute value of the difference between the grayscale value of the j-th pixel (excluding the i-th pixel) and the grayscale value of the i-th pixel.

5. The method for tracking targets in mine roadway inspection based on video recognition according to claim 1, characterized in that, The method for obtaining the illumination uniformity of each pixel based on the noise level of each pixel in the grayscale inspection image, combined with the gradient value of the grayscale value of each pixel with other pixels in its initial matrix window and the local brightness index, includes the following specific methods: ； In the formula, Indicates the first Light uniformity of each pixel Indicates the first The noise level of each pixel This represents the number of pixels in the initial matrix window minus 1. Indicates the first The local brightness index of the j-th pixel (excluding the i-th pixel) within the initial matrix window containing the pixels. Indicates the first The gradient value of the grayscale value of the j-th pixel (excluding the i-th pixel) within the initial matrix window containing all pixels. Indicates the first The average local brightness index of all pixels in the initial matrix window except for the i-th pixel.

6. The method for tracking targets in mine roadway inspection based on video recognition according to claim 1, characterized in that, The method for obtaining the adaptive dark channel image window radius for each pixel by presetting the initial dark channel image window radius for each pixel and combining it with the illumination uniformity of each pixel includes the following specific methods: The initial dark channel radius of each pixel is preset to be r pixels. The adaptive dark channel image window radius of each pixel is obtained based on the illumination uniformity of each pixel. The specific method is as follows: ； Indicates the first Adaptive dark channel image window radius size per pixel Indicates the first Lighting uniformity of each pixel.

7. The method for tracking targets in mine roadway inspection based on video recognition according to claim 1, characterized in that, The specific method for calculating the dark channel image based on the adaptive dark channel image window radius of each pixel to complete the dehazing and enhancement of the inspection image, and using the enhanced inspection image to complete the target tracking of the mine roadway inspection, includes the following: Based on the adaptive dark channel image window radius of each pixel, pixels in the inspection image whose Euclidean distance to each pixel is less than its adaptive dark channel image window radius are recorded as pixels in the adaptive dark channel image window corresponding to each pixel. The dark channel image is obtained based on the grayscale value of the pixels in the adaptive dark channel image window corresponding to each pixel, thus completing the dark channel dehazing of the inspection image and obtaining the enhanced inspection image. The enhanced inspection image is then used to complete the target tracking of the mine roadway inspection.

8. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the video recognition-based target tracking method for mine roadway inspection as described in any one of claims 1 to 7.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the video recognition-based target tracking method for mine roadway inspection as described in any one of claims 1 to 7.

10. A target tracking system for mine roadway inspection based on video recognition, characterized in that, The system includes the following modules: The video preprocessing module is used to acquire mine roadway inspection videos, perform frame extraction on the videos, obtain any grayscale inspection image, and set an initial matrix window corresponding to each pixel in the inspection image as the center. The brightness calculation module is used to obtain the local brightness index of each pixel based on the grayscale value of each pixel in the grayscale inspection image and other pixels in its initial matrix window. The noise calculation module is used to obtain the noise level of each pixel based on the gradient value of the grayscale value of each pixel in the grayscale inspection image and other pixels in its initial matrix window. The illumination uniformity calculation module is used to obtain the illumination uniformity of each pixel based on the noise level of each pixel in the grayscale inspection image, combined with the gradient value of the grayscale value of each pixel and other pixels in its initial matrix window and the local brightness index. The dark channel window size calculation module is used to preset the initial dark channel image window radius for each pixel, and combine the illumination uniformity of each pixel to obtain the adaptive dark channel image window radius size for each pixel; The mine roadway inspection target tracking module is used to calculate the dark channel image based on the adaptive dark channel image window radius of each pixel, complete the dehazing and enhancement of the inspection image, and use the enhanced inspection image to complete the mine roadway inspection target tracking.

Images