Textile carding system based on computer vision

Abstract

The present invention relates to a textile combing system based on computer vision, and belongs to the field of image processing technology. The system includes an airflow intensity control system, which includes a controller and an image acquisition device, the image acquisition device is used to collect images of coarse fibers to be combed on a feed conveyor belt, and the controller is used to: perform preset processing on the acquired image to obtain a target edge feature image; calculate the overlap rate of each pixel point in the edge feature image; construct an overlap rate image according to the overlap rate of each pixel point in the edge feature image, and for any pixel point in the overlap rate image: define a window area of ​​a set size with the pixel point as the center pixel point, and calculate the winding degree of the center pixel point; calculate the dredging difficulty of the overlap rate image according to the winding degree of each pixel point in the overlap rate image, and control the output power of the air compressor according to the dredging difficulty of the overlap rate image. The present invention can perform adaptive adjustment according to the difference of the fibers to be combed.

Description

Technical Field

[0001] The invention relates to the technical field of image processing, and in particular to a textile combing system based on computer vision. Background Art

[0002] The main purpose of the textile combing system is to comb the fibers. There are two main types of textile combing systems on the market. One is the textile combing system for fine fibers, such as cotton and wool; the other is the textile combing system for coarse fibers, such as coconut fiber and ramie.

[0003] The existing textile combing system for coarse fibers includes a feeding mechanism, a discharging mechanism and a combing mechanism arranged between the feeding mechanism and the discharging mechanism. The combing mechanism includes a combing roller. A baffle is arranged above the combing roller to form a combing channel. An interface for connecting an air compressor is provided on the baffle. The air compressor conveys high-speed airflow into the conveying channel through the interface. The high-speed airflow enables the coarse fibers to achieve a higher degree of fit with the combing roller, and the combing effect is better.

[0004] The entanglement complexity of the coarse fibers to be combed is different, and the corresponding combing difficulty is different. If different coarse fibers to be combed are combed indiscriminately according to the same airflow intensity, then when the airflow intensity is small and the combing difficulty of the coarse fibers to be combed is large, the combing effect is difficult to guarantee. When the airflow intensity is large and the combing difficulty of the coarse fibers to be combed is small, unnecessary waste will be caused. Summary of the invention

[0005] In order to solve the problem that the existing textile combing system cannot be adaptively adjusted according to the different fibers to be combed, the present invention provides a technical solution of a textile combing system based on computer vision, including a feed conveyor belt and an air compressor, the textile combing system also includes an airflow intensity control system, the airflow intensity control system includes a controller and an image acquisition device connected to the controller, the image acquisition device is used to collect images of coarse fibers to be combed on the feed conveyor belt, and the controller is used to:

[0006] Performing preset processing on the acquired image of the coarse fiber to be combed, and obtaining a target edge feature image after the processing;

[0007] Calculate the overlap rate of each pixel in the edge feature image: for edge pixels in the edge feature image, calculate the overlap rate of each edge pixel according to the number of fibers passing through each edge pixel and the angle of the extension direction of the fibers passing through each edge pixel; for non-edge pixels in the edge feature image, determine that the overlap rate of each non-edge pixel is 0;

[0008] An overlap ratio image is constructed according to the overlap ratio of each pixel in the edge feature image. For any pixel in the overlap ratio image: a window area of ​​a set size is defined with the pixel as the center pixel, and the winding degree of the center pixel is calculated according to the overlap ratio of each pixel in the area and the distance between each pixel and the center pixel;

[0009] The dredging difficulty of the overlap rate image is calculated according to the entanglement degree of each pixel point in the overlap rate image, and the output power of the air compressor is controlled according to the dredging difficulty of the overlap rate image.

[0010] Beneficial effects: The textile combing system of the present invention also includes an airflow intensity control system, which calculates the difficulty of untangling the coarse fibers to be combed based on the acquired images of the coarse fibers to be combed, and adjusts the output power of the air compressor according to the difficulty of untangling. It can perform adaptive adjustment according to the different fibers to be combed. For coarse fibers to be combed that are more difficult to comb, the output power of the air compressor can be increased to ensure the combing effect; for coarse fibers to be combed that are less difficult to comb, the output power of the air compressor can be reduced to achieve the purpose of saving energy.

[0011] Furthermore, for edge pixels in the edge feature image, the overlap rate of each edge pixel is calculated according to the number of fibers passing through each edge pixel and the angle of the extension direction of the fibers passing through each edge pixel, including:

[0012] Construct a direction chain code, and obtain the extension direction angle of the fiber passing through each edge pixel point according to the direction chain code;

[0013] The overlap rate of each edge pixel is calculated using the following formula:

[0014]

[0015] Among them, C i is the overlap rate of the i-th edge pixel, θ i，j1 represents the angle value of the extension direction of one side of the jth fiber passing through the i-th edge pixel, θ i，j2 represents the angle value of the other side extension direction of the jth fiber passing through the i-th edge pixel point, and N is the number of fibers passing through the i-th edge pixel point.

[0016] Furthermore, for any pixel point in the overlap rate image: a window area of ​​a set size is defined with the pixel point as the center pixel point, and the winding degree of the center pixel point is calculated according to the overlap rate of each pixel point in the area and the distance between each pixel point and the center pixel point, including:

[0017] Calculate the weight of the overlap rate of each pixel in the window area according to the distance between each pixel in the window area and the central pixel;

[0018] The winding degree of the central pixel point of the window area is calculated according to the overlapping ratio of each pixel point in the window area and the weight of the overlapping ratio corresponding to each pixel point;

[0019] The formula for calculating the weight of the overlap rate of each pixel in the window area is:

[0020]

[0021] Among them, d i，z It represents the distance between the zth pixel and the center pixel in the window centered on the i-th pixel in the overlap rate image, Q i，z represents the weight of the zth pixel in the window centered on the i-th pixel in the overlap rate image, and P is the set of pixels in the window centered on the i-th pixel in the overlap rate image.

[0022] Furthermore, the window area of ​​the set size is The process of calculating S includes:

[0023] Obtain the mode of overlap ratios of all pixel points corresponding to edge pixel points in the overlap ratio image;

[0024] The overlap rate image is clustered using a density clustering algorithm to obtain several category sets;

[0025] Calculate the mean in each category set, obtain the category set whose mean is greater than the mode of the overlap rate, and find the mean of the outer rectangular area of ​​these category sets, denoted as S.

[0026] Further, the calculating the entanglement degree of the overlap rate image according to the entanglement degree of each pixel point in the overlap rate image includes:

[0027] The overlap rate image is evenly divided into several regions. For any region: the average of the entanglement degree of each pixel point in the region is calculated according to the entanglement degree of each pixel point in the region, and the average is recorded as the entanglement degree of the region; the influence rate of the upper region is calculated according to the entanglement degree of the upper region of the region; the sum of the entanglement degree of the region and the influence rate of the upper region is taken as the dredging difficulty of the region;

[0028] The average of the dredging difficulties of each region in the overlap ratio image is taken as the dredging difficulty of the overlap ratio image.

[0029] Furthermore, the calculation formula for calculating the upper area influence rate is:

[0030]

[0031] Among them, Y m It represents the influence rate of the upper area of ​​the mth area on the winding degree of the mth area, Tk It represents the degree of entanglement of the area directly above the m-th area, which is k areas away from the m-th area. L represents the average combing loss rate during the fiber combing process. f() represents a positive function. K represents the number of areas directly above the m-th area.

[0032] Further, the controlling the output power of the air compressor according to the dredging difficulty of the overlap rate image includes:

[0033] The dredging difficulty of the overlap rate image is compared with the dredging difficulty threshold range. If the dredging difficulty of the overlap rate image is greater than the upper limit value of the dredging difficulty threshold range, the output power of the air compressor is controlled to be greater than the standard output power; if the dredging difficulty of the overlap rate image is within the dredging difficulty threshold range, the output power of the air compressor is controlled to be the standard output power; if the dredging difficulty of the calculated image is less than the lower limit value of the dredging difficulty threshold range, the output power of the air compressor is controlled to be less than the standard output power.

[0034] Furthermore, the preset processing includes target pixel segmentation, grayscale and edge detection processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is a control principle diagram of the textile combing system of the present invention;

[0036] Figure 2 is a schematic diagram of a control method of a controller of the present invention;

[0037] Figure 3 It is a schematic diagram of the directional chain code matrix of the present invention. DETAILED DESCRIPTION

[0038] To make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be introduced below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0039] The computer vision-based textile combing system of this embodiment includes a feeding mechanism, a discharging mechanism, and a combing mechanism arranged between the feeding mechanism and the discharging mechanism. The combing mechanism includes a combing roller, a baffle is arranged above the combing roller and forms a combing channel, and an interface for connecting an air compressor is arranged on the baffle, and the air compressor conveys a high-speed airflow into the conveying channel through the interface; on this basis, Figure 1 As shown, the computer vision-based textile combing system of this embodiment also includes an airflow intensity control system, such as Figure 1As shown in the dotted box, the airflow intensity control system includes a controller and an image acquisition device connected to the controller. The image acquisition device is arranged above the feed conveyor belt, and is used to acquire images of the coarse fibers to be combed on the feed conveyor belt; the image acquisition device transmits the acquired images to the controller, and the controller analyzes the combing difficulty of the coarse fibers to be combed according to the acquired images, and controls the output power of the air compressor according to the analysis results, so as to change the intensity of the airflow delivered by the air compressor according to the combing difficulty of the coarse fibers.

[0040] The computer vision-based textile combing system of this embodiment is an existing textile combing system with an additional airflow intensity control system. In view of the fact that the structure, working principle and working process of the existing textile combing system are well-known technologies, this embodiment will no longer elaborate on the structure, working principle and working process of the computer vision-based textile combing system of this embodiment, but will focus on how the controller controls the output power of the air compressor based on the image captured by the image acquisition device.

[0041] like Figure 2 As shown, the controller executes the following control method according to the received image of the coarse fiber to be combed:

[0042] (1) performing target pixel segmentation, grayscale conversion and edge detection processing on the acquired image of the coarse fiber to be combed to obtain a target edge feature image;

[0043] The image acquisition device of this embodiment is a high-definition camera, which looks down and acquires images of the coarse fibers to be combed at a preset frequency, and the acquired images of the coarse fibers to be combed are RGB images. This embodiment performs relevant processing on the acquired RGB images, and the specific process is as follows:

[0044] Target pixel segmentation processing: The target pixel of this embodiment is the coarse fiber pixel. The process of segmenting the coarse fiber pixel in the image in this embodiment is as follows: the color interval of the RGB channel of the coarse fiber to be combed is determined according to the color empirical value of the coarse fiber to be combed [a R , b R ],[a G , b G ],[a B , b B ], where a R is the lower limit of the red channel of crude fiber, b R is the upper limit of the red channel of crude fiber, a G is the lower limit of the green color channel of crude fiber, b G is the upper limit of the green color channel of crude fiber, a B is the lower limit of the blue color channel of crude fiber, b Bis the upper limit of the blue color channel of the coarse fiber; separate the color values ​​of each channel h1 x，y ∈[a R , b R ], h2 x，y ∈[a G , b G ], h3 x，y ∈[a B , b B ] pixel point set, determine the pixel points in the pixel point set as coarse fiber pixel points, h1 x，y is the red color channel value of the pixel (x, y), h2 x，y h3 is the green channel value of the pixel (x, y), x，y is the blue color channel value of the pixel point (x, y), x is the horizontal coordinate of the pixel point, and y is the vertical coordinate of the pixel point; the pixels in the image except the coarse fiber pixels are determined as background pixels, and the color values ​​of each channel of the background pixels are set to zero to obtain the processed image T0.

[0045] Grayscale processing: convert image T0 from an RGB color image to a grayscale image T1, where grayscale image T1 is the target grayscale image. The grayscale processing process is prior art and will not be described in detail here.

[0046] Edge detection processing: Use the Canny algorithm to perform edge detection on the T1 image to obtain an edge feature image, that is, a target edge feature image. The edge detection processing process is an existing technology and will not be described in detail here.

[0047] (2) Calculating the overlap rate of each pixel in the edge feature image: For edge pixels in the edge feature image, the overlap rate of each edge pixel is calculated according to the number of fibers passing through each edge pixel and the angle of the extension direction of the fibers passing through each edge pixel; for non-edge pixels in the edge feature image, the overlap rate of each non-edge pixel is determined to be 0;

[0048] The pixels in the edge feature image are divided into two categories, one is edge pixels and the other is non-edge pixels; for non-edge pixels, the number of fibers passing through them is 0. This embodiment determines that the overlap rate of non-edge pixels is 0. The overlap rate of edge pixels is related to the number of fibers passing through the edge pixels and the extension direction angle of the coarse fibers passing through the edge pixels. The process of calculating the overlap rate of each edge pixel in this embodiment is as follows:

[0049] ① Construct the direction chain code: the vertical direction is 0 degrees, the horizontal direction is 90 degrees, and the direction chain code matrix is ​​as follows: Figure 3 As shown; in this embodiment, the 0 degree direction corresponds to the combing direction of the textile combing system.

[0050] ②According to the direction chain code, obtain the extension direction angle:

[0051] For any edge pixel in the edge feature image, take it as the center pixel and obtain the angle θ of all fiber extension directions passing through the center pixel. i，j1 ,θ i，j2 ,θ i，j1 represents the angle value of the extension direction of one side of the jth fiber passing through the i-th edge pixel, θ i，i2 Represents the angle value of the other side extension direction of the jth fiber passing through the i-th edge pixel.

[0052] The angles of the two extending directions of all fibers passing through the ith edge pixel point can be obtained in the above manner. The method for determining the number of fibers passing through the ith edge pixel point in this embodiment is as follows: by performing connectivity analysis, the number of all edges connected to the ith edge pixel point is obtained, and the number of edges is divided by 2 to obtain the number of fibers passing through the ith pixel point.

[0053] ③According to the angle of the fiber extension direction, the overlap rate is obtained:

[0054]

[0055] Among them, C i is the overlap rate of the i-th edge pixel, θ i，j1 represents the angle value of the extension direction of one side of the jth fiber passing through the i-th edge pixel, θ i，j2 Indicates the angle value of the other side extension direction of the jth fiber passing through the i-th edge pixel. 1-cos(θ i，j1 ) represents the redundant fiber length on one side of the jth fiber passing through the ith edge pixel, 1-cos(θ i，j2 ) represents the redundant length of the other side fiber of the jth fiber passing through the i-th edge pixel. (1-cos(θ i，j1 )+(1-cos(θ i，j2 ) represents the fiber length redundancy of the jth fiber passing through the i-th edge pixel, It represents the total fiber length redundancy passing through the i-th edge pixel, that is, the fiber overlap rate at the i-th edge pixel, and N is the number of fibers passing through the i-th edge pixel.

[0056] (3) constructing an overlap ratio image according to the overlap ratio of each pixel in the edge feature image. For any pixel in the overlap ratio image, a window area of ​​a set size is defined with the pixel as the center pixel, and the winding degree of the center pixel is calculated according to the overlap ratio of each pixel in the area and the distance between each pixel and the center pixel.

[0057] I. Construct an overlap ratio image with the overlap ratio as the pixel value, and obtain the mode of the overlap ratios of all pixel points corresponding to the edge pixels in the overlap ratio image.

[0058] II. Use the density clustering algorithm to perform density clustering on the overlap rate image to obtain several category sets. The overlap rates of the pixels in each category set are relatively close. Calculate the mean in each category set to obtain the category sets whose means are greater than the mode of the overlap rate, and find the outer rectangular areas of these category sets. Find the mean of these rectangular areas as the general area S for judging the fiber entanglement.

[0059] III. For the overlap ratio image, is the window size, and the weight of the overlap rate of each pixel in the window is calculated:

[0060]

[0061] Among them, d i，z It represents the distance between the zth pixel and the center pixel in the window centered on the i-th pixel in the overlap rate image, Q i，z represents the weight of the zth pixel in the window centered on the i-th pixel in the overlap rate image, and P is the set of pixels in the window centered on the i-th pixel in the overlap rate image.

[0062] IV. After obtaining the weights of each pixel in the window, use the weighted summation formula to calculate the mean overlap rate of the corresponding area of ​​the window, and use the mean as the winding degree R of the central pixel of the window. i .

[0063]

[0064] Among them, R i Indicates the degree of entanglement of the i-th pixel in the overlap rate image, C i，z Represents the overlap rate of the zth pixel in the window centered at the i-th pixel in the overlap rate image.

[0065] (4) The entanglement degree of the overlap ratio image is calculated according to the entanglement degree of each pixel point in the overlap ratio image, and the output power of the air compressor is controlled according to the entanglement degree of the overlap ratio image.

[0066] Divide the overlap rate image evenly into several For any region: the mean entanglement degree of each pixel point in the region is calculated according to the entanglement degree of each pixel point in the region, and the mean is recorded as the entanglement degree of the region.

[0067] Since the accumulated fibers in the upper area will be transferred to the lower side after combing on the upper side, which will affect the combing difficulty of the lower area, the present embodiment also analyzes the influence rate of the fiber entanglement degree in the upper area on the entanglement degree in the lower area; considering that a part of the accumulated fibers on the upper side will be directly combed and fallen off during the process of combing to the lower area, the influence of such short fibers on the lower side is relatively small, and the present embodiment calculates the influence rate of the upper area based on the following formula.

[0068]

[0069] Among them, Y m It represents the influence rate of the upper area of ​​the mth area on the winding degree of the mth area, T k It represents the degree of entanglement of the area directly above the mth area, which is k areas away from the mth area. L represents the average combing loss rate during the fiber combing process, which is an empirical value. f() represents a positive function, that is, when the independent variable is positive, the original value is retained, and when the independent variable is negative, the function value is zero. represents the dredging influence rate of all the areas directly above the m-th area on the m-th area, and K represents the number of areas directly above the m-th area, which can be determined according to the coordinate position relationship of the central pixel points of each area.

[0070] After obtaining the influence rate Y of the upper area of ​​the mth area on the winding degree of the mth area m Based on the influence rate Y m and the degree of entanglement T of the mth region m Calculate the dredging difficulty of the mth area:

[0071] W m =exp(Y m +T m )

[0072] Among them, T m represents the winding degree of the mth region, W m Indicates the difficulty of dredging the mth area.

[0073] Thus, the overlap ratio in the image can be calculated. The dredging difficulty of the area, the average of these areas is taken as the dredging difficulty of the overlap rate image, and the dredging difficulty of the image is compared with the dredging difficulty threshold range [W 1 , W 2 ] for comparison, W 1 <W 2In this embodiment, the dredging difficulty threshold range corresponds to the standard output power of the air compressor, that is, when the dredging difficulty to be combed is within the dredging difficulty threshold range, controlling the air compressor to output according to the standard output power can enable the coarse fibers to be combed to obtain a combing effect that meets the requirements; the corresponding relationship between the dredging difficulty threshold range and the standard output power of the air compressor in this embodiment can be determined based on experience or through multiple tests, which will not be repeated here.

[0074] If the calculated image dredging difficulty is greater than the upper limit value W of the dredging difficulty threshold range 2 , indicating that the dredging difficulty of the fiber to be dredged is relatively large, the output power of the air compressor is controlled to be greater than the standard output power, so that the airflow intensity output by the air compressor becomes larger, so as to increase the degree of fit between the fiber to be dredged and the combing roller to meet the combing requirements; if the dredging difficulty of the calculated image is within the dredging difficulty threshold range, it means that the dredging difficulty of the fiber to be dredged is the standard difficulty, which is relatively matched with the output power of the air compressor, and the output power of the air compressor is controlled to be the standard output power; if the dredging difficulty of the calculated image is less than the lower limit value W of the dredging difficulty threshold range 1 , indicating that the difficulty of clearing the fibers to be cleared is relatively small, and the output power of the air compressor can be controlled to be lower than the standard output power to save energy.

[0075] The textile combing system of this embodiment also includes an airflow intensity control system, which calculates the difficulty of untangling the coarse fibers to be combed based on the acquired images of the coarse fibers to be combed, and adjusts the output power of the air compressor according to the difficulty of untangling. It can be adaptively adjusted according to the different fibers to be combed. For coarse fibers to be combed that are more difficult to comb, the output power of the air compressor can be increased to ensure the combing effect; for coarse fibers to be combed that are less difficult to comb, the output power of the air compressor can be reduced to achieve the purpose of saving energy.

[0076] It should be noted that, although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

Claims

1. A textile combing system based on computer vision, comprising a feed conveyor belt and an air compressor, characterized in that: The textile combing system further comprises an airflow intensity control system, the airflow intensity control system comprising a controller and an image acquisition device connected to the controller, the image acquisition device being used to acquire images of the coarse fibers to be combed on the feed conveyor belt, and the controller being used to: Performing preset processing on the acquired image of the coarse fiber to be combed, and obtaining a target edge feature image after the processing; Calculate the overlap rate of each pixel in the edge feature image: for edge pixels in the edge feature image, calculate the overlap rate of each edge pixel according to the number of fibers passing through each edge pixel and the angle of the extension direction of the fibers passing through each edge pixel; for non-edge pixels in the edge feature image, determine that the overlap rate of each non-edge pixel is 0; An overlap ratio image is constructed according to the overlap ratio of each pixel in the edge feature image. For any pixel in the overlap ratio image: a window area of ​​a set size is defined with the pixel as the center pixel, and the winding degree of the center pixel is calculated according to the overlap ratio of each pixel in the area and the distance between each pixel and the center pixel; Calculating the dredging difficulty of the overlap rate image according to the entanglement degree of each pixel point in the overlap rate image, and controlling the output power of the air compressor according to the dredging difficulty of the overlap rate image; For edge pixels in the edge feature image, the overlap rate of each edge pixel is calculated according to the number of fibers passing through each edge pixel and the angle of the extension direction of the fibers passing through each edge pixel, including: Construct a direction chain code, and obtain the extension direction angle of the fiber passing through each edge pixel point according to the direction chain code; The overlap rate of each edge pixel is calculated using the following formula: in, is the overlap rate of the i-th edge pixel, It represents the angle value of the extension direction of one side of the j-th fiber passing through the i-th edge pixel point. represents the angle value of the other side extension direction of the jth fiber passing through the i-th edge pixel point, and N is the number of fibers passing through the i-th edge pixel point; For any pixel point in the overlap rate image: a window area of ​​a set size is defined with the pixel point as the center pixel point, and the winding degree of the center pixel point is calculated according to the overlap rate of each pixel point in the area and the distance between each pixel point and the center pixel point, including: Calculate the weight of the overlap rate of each pixel in the window area according to the distance between each pixel in the window area and the central pixel; The winding degree of the central pixel point of the window area is calculated according to the overlapping ratio of each pixel point in the window area and the weight of the overlapping ratio corresponding to each pixel point; The formula for calculating the weight of the overlap rate of each pixel in the window area is: in, It represents the distance between the zth pixel point and the center pixel point in the window centered on the i-th pixel point in the overlap rate image. represents the weight of the z-th pixel in the window centered on the i-th pixel in the overlap rate image, and P is the set of pixels in the window centered on the i-th pixel in the overlap rate image; The step of calculating the winding degree of the overlap rate image according to the winding degree of each pixel point in the overlap rate image comprises: The overlap rate image is evenly divided into several regions. For any region: the average of the entanglement degree of each pixel point in the region is calculated according to the entanglement degree of each pixel point in the region, and the average is recorded as the entanglement degree of the region; the influence rate of the upper region is calculated according to the entanglement degree of the upper region of the region; the sum of the entanglement degree of the region and the influence rate of the upper region is taken as the dredging difficulty of the region; The average of the dredging difficulties of each area in the overlap rate image is taken as the dredging difficulty of the overlap rate image; The calculation formula for calculating the upper area influence rate is: in, It represents the influence rate of the upper area of ​​the mth area on the winding degree of the mth area, represents the degree of winding of the region directly above the mth region, which is k regions away. represents the average combing loss rate during fiber combing, () represents a positive function, Indicates the number of regions directly above the mth region.

2. The computer vision-based textile combing system according to claim 1, characterized in that: The window area of ​​the set size is The window area is calculated The process includes: Obtain the mode of overlap ratios of all pixel points corresponding to edge pixel points in the overlap ratio image; The overlap rate image is clustered using a density clustering algorithm to obtain several category sets; Calculate the mean in each category set, obtain the category set whose mean is greater than the mode of the overlap rate, and find the mean of the outer rectangular area of ​​these category sets, denoted as S.

3. The computer vision-based textile combing system according to claim 1, characterized in that: The method of controlling the output power of the air compressor according to the dredging difficulty of the overlap rate image includes: The dredging difficulty of the overlap rate image is compared with the dredging difficulty threshold range. If the dredging difficulty of the overlap rate image is greater than the upper limit value of the dredging difficulty threshold range, the output power of the air compressor is controlled to be greater than the standard output power; if the dredging difficulty of the overlap rate image is within the dredging difficulty threshold range, the output power of the air compressor is controlled to be the standard output power; if the dredging difficulty of the calculated image is less than the lower limit value of the dredging difficulty threshold range, the output power of the air compressor is controlled to be less than the standard output power.

4. The computer vision-based textile combing system according to claim 1, characterized in that: The preset processing includes target pixel segmentation, grayscale and edge detection processing.

Images