A method for calculating the surface coating characteristic index of sizing yarn based on image processing technology and a device for collecting sizing yarn dryness sequence images

Abstract

The present invention relates to a method for calculating the surface coating characteristic index of sizing yarn based on image processing technology and a device for collecting sizing yarn strip sequence images. The method includes collecting sizing yarn strip sequence images, processing the images, obtaining sizing yarn effective statistical area images, using the intersection of the sizing yarn pixel gray value normal distribution curve and the original yarn pixel gray value normal distribution curve as the threshold T for covering area segmentation, and finally calculating the sizing yarn coating coefficient SCC, coating depth SCD and coating unevenness SCU by formula, and using these three indicators to evaluate the sizing quality of sizing yarn. The device includes a solution tank, and unwinding devices and winding devices are respectively arranged at both ends of the solution tank, a shooting device is arranged above the solution tank, and a waste liquid tank is also arranged between the solution tank and the winding device, and a pair of rollers are arranged on the waste liquid tank.

Description

Technical Field

[0001] The invention belongs to the technical field of textile sizing, and in particular relates to a method for calculating surface coating characteristic indexes of sizing based on image processing technology and a device for collecting sizing sequence images. Background Art

[0002] The coating penetration characteristics of sizing yarn are important indicators for evaluating the quality of sizing yarn. Sizing coating is beneficial to improving the wear resistance of sizing yarn, making the hairiness flat and the yarn surface smooth; the penetration of sizing material helps the bonding between fibers, thereby improving the strength of sizing yarn. The unreasonable distribution of sizing yarn coating penetration will cause the reduction of sizing yarn performance, and the instability of sizing yarn coating penetration will also cause fluctuations in sizing yarn performance. Therefore, the reasonable configuration and stability of sizing yarn coating penetration are the key factors to improve sizing yarn performance and constant sizing quality.

[0003] At present, the slicing method is often used to evaluate the coating of sizing yarn. The hand-drawn slicing method is to obtain the sizing yarn slices through a slicer, take the slice images based on a microscope, and use homogeneous paper to manually draw the relevant perimeter of the sizing yarn cross section, so as to calculate the sizing yarn coating characteristics by the area calculation method or the weighing method. On this basis, some scholars applied computer image technology to the processing of slices and proposed the image slicing method, which effectively improved the efficiency of the slicing method to test the coating characteristics of sizing yarn. However, the slicing method has certain limitations in characterizing the coating of sizing yarn. First, the thickness of the sizing yarn segment cut by the slicing method is too thin to represent the quality of the entire sizing yarn, and the integrity of the sizing film on the surface of the sizing yarn is easily destroyed during the operation, and the repeatability is poor. Secondly, for the cross-sectional images of sizing yarn taken by a microscope, the boundary between the sizing yarn coating and the penetration area is unclear, which is easy to cause large errors. Thirdly, the slicing method extracts the sizing yarn coating characteristic indicators of the sizing liquid coating rate and the sizing film integrity rate. However, the high or low sizing liquid coating rate cannot reflect the quality of the coating. Therefore, it is impossible to provide reverse guidance for the performance and sizing process of the sizing yarn. Summary of the invention

[0004] The first technical problem to be solved by the present invention is to provide a method for calculating the surface coating characteristic index of the sizing yarn based on image processing technology to accurately calculate the surface coating coefficient, coating depth and coating unevenness of the sizing yarn, so as to accurately evaluate the surface coating effect of the sizing yarn.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is: a method for calculating the surface coating characteristic index of sizing yarn based on image processing technology, comprising the following steps:

[0006] (1) coloring the sizing film of the sizing yarn;

[0007] (2) photographing the sizing yarn in sequence along its length direction to obtain a certain number of sizing yarn strip sequence images;

[0008] (3) graying, noise reduction, background removal, and tilt correction are performed on the above-mentioned sizing yarn sequence images in turn by computer image processing technology to obtain a certain number of sizing yarn sequence grayscale images;

[0009] (4) extracting an image of the effective statistical area of ​​the sizing yarn from each grayscale image of the sizing yarn sequence;

[0010] (5) stitching a certain number of effective statistical area images of sizing yarn to form a long segment image of sizing yarn;

[0011] (6) processing the original yarn by using the above steps (2) to (5) to obtain a long segment image of the original yarn;

[0012] (7) making a grayscale distribution histogram of all pixels in the long segment image of the sizing yarn and performing normal distribution fitting to obtain a normal distribution curve of grayscale values ​​of pixels in the sizing yarn; making a grayscale distribution histogram of all pixels in the long segment image of the original yarn and performing normal distribution fitting to obtain a normal distribution curve of grayscale values ​​of pixels in the original yarn; taking the intersection of the normal distribution curve of grayscale values ​​of pixels in the sizing yarn and the normal distribution curve of grayscale values ​​of pixels in the original yarn as the threshold T for segmenting the covered area, and the area in the grayscale distribution histogram of the sizing yarn where the grayscale value is lower than the threshold T is the area where the color reaction occurs, that is, the covered area;

[0013] (8) Through the formula The grayscale distribution histogram of the effective statistical area of ​​the sizing yarn and the original yarn is converted from normal distribution to standard normal distribution; where Y represents the threshold in the standard normal distribution, T represents the normal distribution threshold, represents the average gray value of the pixel points in the sizing long segment image, and σ represents the standard deviation of the gray value in the sizing long segment image; then, by looking up the "standard normal distribution table", the probability value less than the threshold value Y is obtained, that is, the sizing coverage coefficient SCC;

[0014] (9) Through the formula Calculate the coating depth and evaluate the coating thickness of the sizing yarn with the coating depth SCD; where G i Represents the gray value of the i-th pixel in the sizing long segment image, N y Indicates the number of pixels in the sizing long segment image;

[0015] (10) Through the formula Calculate the coating unevenness and evaluate the coating unevenness of the sizing yarn with the coating unevenness SCU; where G i represents the gray value of the i-th pixel in the sizing long segment image, It represents the average gray value of pixels in the long segment image of sizing yarn.

[0016] As a preferred solution, the specific method for extracting the effective statistical area image of the sizing yarn in step (4) is: respectively calculate the average grayscale value of each row of pixels in each grayscale image of the sizing yarn sequence, and select the area where the difference between the maximum average grayscale value and the minimum average grayscale value is less than 2 and the number of consecutive rows is greater than 10 as the effective statistical area image of the sizing yarn.

[0017] As a preferred solution, in step (3), the specific method of noise reduction processing is median filtering denoising; the specific method of background removal is: binarization is used to segment the image, the background other than the sizing yarn is removed, and the sizing yarn dryness image is obtained; the specific method of tilt correction is: the image is tilted and corrected by Hough transform.

[0018] As a preferred solution, in the verification step (7), the grayscale distribution of all pixels in the sizing long segment image conforms to the normal distribution, and the grayscale distribution of all pixels in the original yarn long segment image conforms to the normal distribution.

[0019] As a preferred solution, the number of the sizing yarn sequence images is not less than 40.

[0020] The beneficial effects of the present invention are as follows: the present invention takes the longitudinal appearance image of the sizing yarn (i.e., the sizing yarn strip sequence image) as the object, and obtains the sizing yarn strip sequence grayscale image through image graying, median filtering denoising, binarization background removal, Hough transform tilt correction and other image preprocessing. Based on the setting and optimization of parameters such as the coating area threshold, image acquisition sample capacity and the effective statistical area of ​​the sizing yarn strip sequence image, the coating coefficient SCC, coating depth SCD and coating unevenness SCU are calculated by formula to evaluate the coating effect of the sizing yarn surface. The coating scale of the sizing yarn surface is very small and cannot be accurately characterized by relevant macroscopic testing methods, while image processing technology is efficient, simple and accurate, and has a high sensitivity to subtle coating changes on the sizing yarn surface. The calculation of the coating characteristic index of the sizing yarn can reversely guide the production process of the sizing yarn, thereby improving the sizing yarn performance and stabilizing the sizing quality, which has broad practical production significance.

[0021] The present invention further analyzes the grayscale values ​​of each row of pixel points in the sizing yarn area, obtains the number of pixel rows with relatively stable grayscale values ​​as the effective statistical area, and is used to calculate the characteristic index of the sizing yarn surface coating. By determining the effective statistical area, the influence of lighting factors on the sizing yarn shooting image is avoided, and the evaluation accuracy of the sizing yarn coating is improved.

[0022] The present invention further splices a certain number of collected sequence images with effective statistical areas into long segments, takes the segment images as objects, calculates the sizing yarn coverage coefficient, coverage depth and coverage unevenness of the segment length, expands the sample area, and is more representative for the test of sizing yarn coverage.

[0023] The present invention further dynamically sets the threshold of the color development area. The grayscale distribution histogram in the effective statistical area of ​​the sizing yarn and the original yarn is obtained, and the intersection of the normal distribution fitting curves of the two histograms is set as the threshold. The area in the sizing yarn with a grayscale value lower than the threshold is the covered area. By dynamically setting the threshold of the covered area, the interference of the fixed threshold on the determination of the covered area with different sizing rates is avoided, and the covered area is obtained more accurately.

[0024] The present invention further studies the image acquisition sample capacity, obtains the minimum sample capacity value that can meet the stable test results, and determines the sizing image acquisition sample capacity, so as to achieve the purpose of stable, accurate and efficient sizing coating evaluation.

[0025] A further technical problem to be solved by the present invention is to provide a safe and environmentally friendly device for collecting sizing yarn evenness sequence images.

[0026] In order to solve the above technical problems, the technical solution adopted by the present invention is: a device for collecting sequential images of sizing yarns, comprising a solution tank, with an unwinding device and a winding device respectively provided at both ends of the solution tank, a shooting device is provided above the solution tank, the end of the solution tank close to the unwinding device is the upstream end, and the end close to the winding device is the downstream end, a waste liquid tank is also provided between the solution tank and the winding device, and a pair of rollers are provided on the waste liquid tank.

[0027] As a preferred embodiment, the unwinding device includes a plurality of unwinding tubes which are vertically arranged and arranged perpendicular to the moving direction of the slurry yarn, an L-shaped support plate is arranged at one end of the arrangement direction of the unwinding tubes, the upper end of the support plate is higher than the top end of the unwinding tube and is connected to a horizontally arranged first guide roller, the axial direction of the first guide roller is parallel to the arrangement direction of each unwinding tube, the first guide roller is located directly above each unwinding tube, and the lower part of the support plate is connected to a second guide roller which is parallel to the first guide roller, and the second guide roller is located between the unwinding tube and the solution tank; the winding device includes shaft seats which are arranged on both sides of the solution tank and a winding shaft which is rotatably connected between the two shaft seats, a motor is connected to the outer side of one of the shaft seats, the motor is transmission-connected to the winding shaft, and drives the winding shaft to rotate; two combing reeds are arranged in the solution tank, and the shooting device is located between the two combing reeds.

[0028] As a preferred solution, an upstream tensioner for applying tension to the sizing yarn is provided between the unwinding device and the solution tank, and a downstream tensioner for applying tension to the sizing yarn is provided between the waste liquid tank and the winding device.

[0029] As a preferred solution, the shooting device includes a high-speed camera and two strip light sources arranged upstream and downstream of the high-speed camera. The high-speed camera and the two strip light sources are installed on a photography frame, which is fixed on one side of the solution tank.

[0030] The invention provides a waste liquid tank and a pair of rollers to squeeze out the developed sizing yarn and receive the waste liquid, thereby preventing the waste liquid from contaminating the winding shaft when the sizing yarn is wound onto the winding shaft, thereby improving the safety and environmental protection performance of the device.

[0031] The present invention further realizes the simultaneous acquisition of sequence images of multiple sizing yarns by setting multiple unwinding tubes, thereby improving the detection efficiency, and at the same time improving the consistency of the shooting environment and the accuracy of the final detection result. At the same time, the present invention uses a carding reed set in the solution tank as a guide for the sizing yarn in the solution tank, further improving the independence of the sizing yarn operation, avoiding the entanglement of adjacent sizing yarns in the solution tank, and ensuring that the collected sequence images of each sizing yarn do not interfere with each other.

[0032] The present invention further controls the tension of the sizing yarn in the solution tank by arranging an upstream tensioner and a downstream tensioner, so that the tension of the sizing yarn is kept stable during the traction process, thereby reducing or even eliminating the deviation between the overall length of the collected sizing yarn sequence image and the theoretical length, thereby improving the final detection accuracy.

[0033] The present invention further arranges a strip light source upstream and downstream of the high-speed camera respectively to ensure the light intensity required for shooting. At the same time, the use of the strip light source can make the lighting effect of the yarn at various positions in the axial direction tend to be consistent, thereby improving the accuracy of the grayscale value of the pixel points on the yarn surface in each sizing yarn sequence image and improving the final detection accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] The specific embodiments of the present invention are further described in detail below with reference to the accompanying drawings, wherein:

[0035] Figure 1 It is the grayscale histogram fitting curve of 15 rows of pixels in the central axis area of ​​the sized yarn and the original yarn;

[0036] Figure 2 It is a schematic diagram of the structure of a device for collecting a sequence of images of sizing yarn;

[0037] Figure 2 Middle: 1. solution tank, 2. unwinding device, 201. unwinding tube, 202. support plate, 203. first guide roller, 204. second guide roller, 3. winding device, 301. shaft seat, 302. winding shaft, 303. motor, 4. shooting device, 401. high-speed camera, 402. strip light source, 403. photography stand, 5. waste liquid tank, 6. roller, 7. sizing, 8. upstream tensioner, 9. downstream tensioner, 10. combing reed, 11. guide roller. DETAILED DESCRIPTION

[0038] The present invention is explained in detail below through specific embodiments.

[0039] A method for calculating the surface coating characteristic index of sizing yarn based on image processing technology comprises the following steps:

[0040] (1) Soak the sized yarn in I2-KI solution to allow the sized yarn film to develop color.

[0041] (2) Use a camera to take pictures along the length direction of the sizing yarn to obtain 60 sequential images of the sizing yarn.

[0042] (3) The above-mentioned sizing yarn dryness sequence images are grayed, denoised, background removed, and tilted corrected in turn by computer image processing technology to obtain 60 sizing yarn dryness sequence grayscale images. The specific method of denoising is median filtering denoising; the specific method of background removal is: binarization to segment the image, remove the background other than the sizing yarn, and obtain the sizing yarn dryness image; the specific method of tilt correction is: tilt detection and correction of the image by Hough transform.

[0043] (4) Calculate the average grayscale value of each row of pixels in each grayscale image of the sizing yarn sequence, and select the area where the difference between the maximum average grayscale value and the minimum average grayscale value is less than 2 and the number of consecutive rows is greater than 10 as the effective statistical area image of the sizing yarn. In this embodiment, it is found through statistics that the grayscale value is relatively stable within the range of 15 consecutive rows on the central axis and both sides of the sizing yarn. Therefore, 15 consecutive rows of pixels near the central axis of each grayscale image of the sizing yarn sequence are intercepted as the effective statistical area image.

[0044] (5) stitching all the effective statistical area images of the sizing yarn to form a long segment image of the sizing yarn;

[0045] (6) Processing the raw yarn by the method described in the above steps (2) to (5) to obtain a long-segment image of the raw yarn; specifically: using a camera to sequentially shoot along the length direction of the raw yarn to obtain 60 raw yarn strip sequence images; using computer image processing technology to sequentially perform grayscale processing, noise reduction, background removal, and tilt correction on the above raw yarn strip sequence images to obtain 60 raw yarn strip sequence grayscale images; calculating the average grayscale value of each row of pixels in each raw yarn strip sequence grayscale image, and selecting an area where the difference between the maximum average grayscale value and the minimum average grayscale value is less than 2 and the number of consecutive rows is greater than 10 as the raw yarn effective statistical area image. In this embodiment, it is found through statistics that the grayscale value is relatively stable within the range of 15 consecutive rows on the central axis and both sides of the raw yarn. Therefore, 15 consecutive rows of pixels near the central axis of each raw yarn strip sequence grayscale image are intercepted as the effective statistical area image, and all the raw yarn effective statistical area images are image stitched to form the raw yarn long-segment image;

[0046] (7) Figure 1As shown, a grayscale distribution histogram of all pixels in the sizing long segment image is made and normal distribution fitting is performed to obtain a normal distribution curve of grayscale values ​​of sizing pixels; a grayscale distribution histogram of all pixels in the original yarn long segment image is made and normal distribution fitting is performed to obtain a normal distribution curve of grayscale values ​​of original yarn pixels; the intersection of the grayscale value normal distribution curve of sizing pixels and the grayscale value normal distribution curve of original yarn pixels is used as the threshold T for segmenting the covered area, and the area in the grayscale distribution histogram of the sizing grayscale value is lower than the threshold T as the area where the color reaction occurs, that is, the covered area.

[0047] In the verification step (7), the grayscale distribution of all pixels in the sizing long segment image conforms to the normal distribution, and the grayscale distribution of all pixels in the original yarn long segment image conforms to the normal distribution.

[0048] (8) Through the formula The grayscale distribution histogram of the effective statistical area of ​​the sizing yarn and the original yarn is converted from normal distribution to standard normal distribution; where Y represents the threshold in the standard normal distribution, T represents the normal distribution threshold, represents the average gray value of the pixel points in the sizing long segment image, and σ represents the standard deviation of the gray value in the sizing long segment image; then, by looking up the "standard normal distribution table", the probability value less than the threshold value Y is obtained, that is, the sizing coverage coefficient SCC;

[0049] (9) Through the formula Calculate the coating depth and use the coating depth SCD to evaluate the coating thickness of the sizing yarn; where G i Represents the gray value of the i-th pixel in the sizing long segment image, N y Indicates the number of pixels in the sizing long segment image;

[0050] (10) Through the formula Calculate the coating unevenness and evaluate the coating unevenness of the sizing yarn with the coating unevenness SCU; where G i represents the gray value of the i-th pixel in the sizing long segment image, It represents the average gray value of pixels in the long segment image of sizing yarn.

[0051] In this embodiment, 60 sizing yarn sequence images are obtained. In the actual detection process, through the experience accumulated from a large number of experiments, we believe that the number of sizing yarn sequence images should not be less than 40. When it is less than 40, the evaluation results will be unstable and have large deviations.

[0052] The above method was used to test the 50 S The surface coating characteristic index of the sizing yarn was calculated, and the results are shown in Table 1:

[0053] Table 1:

[0054]

[0055] When implementing the above step (2), a device for collecting sequential images of sizing yarn strips is used to collect sequential images of the strips, the device comprising a solution tank 1, with an unwinding device 2 and a winding device 3 respectively provided at both ends of the solution tank 1, a shooting device 4 is provided above the solution tank 1, the end of the solution tank 1 close to the unwinding device 2 is the upstream end, and the end close to the winding device 3 is the downstream end, a waste liquid tank 5 is also provided between the solution tank 1 and the winding device 3, and a pair of rollers 6 are provided on the waste liquid tank 5.

[0056] The unwinding device 2 includes a plurality of unwinding tubes 201 arranged vertically and perpendicular to the moving direction of the sizing yarn 7, an L-shaped support plate 202 is arranged at one end of the arrangement direction of the unwinding tubes 201, the upper end of the support plate 202 is higher than the top of the unwinding tubes 201 and is connected to a horizontally arranged first yarn guide roller 203, the axial direction of the first yarn guide roller 203 is parallel to the arrangement direction of each unwinding tube 201, the first yarn guide roller 203 is located directly above each unwinding tube 201, and the lower part of the support plate 202 is connected to a first yarn guide roller 203 parallel to the first yarn guide roller 203. The second yarn guide roller 204 of the yarn roller 203 is located between the unwinding tube 201 and the solution tank 1; the winding device 3 includes shaft seats 301 arranged on both sides of the solution tank 1 and a winding shaft 302 rotatably connected between the two shaft seats 301, wherein a motor 303 is connected to the outer side of one of the shaft seats 301, and the motor 303 is transmission-connected to the winding shaft 302 to drive the winding shaft 302 to rotate; two combing reeds 10 are arranged in the solution tank 1, and the shooting device 4 is located between the two combing reeds 10.

[0057] An upstream tensioner 8 for applying tension to the sizing yarn 7 is arranged between the unwinding device 2 and the solution tank 1 , and a downstream tensioner 9 for applying tension to the sizing yarn 7 is arranged between the waste liquid tank 5 and the winding device 3 .

[0058] The photographing device 4 includes a high-speed camera 401 and two strip light sources 402 respectively arranged upstream and downstream of the high-speed camera 401 . The high-speed camera 401 and the two strip light sources 402 are both mounted on a photographing frame 403 , which is fixedly arranged on one side of the solution tank 1 .

[0059] When in use, I2-KI solution is injected into the solution tank 1, and then the yarn tube 12 wound with the sizing yarn is inserted into the unwinding tube 201, and the sizing yarn is guided to bypass the first guide roller 203 and the second guide roller 204, and then guided into the I2-KI solution in the solution tank 1 through the upstream tensioner 8 and the two guide rollers 11 located at the upstream end of the solution tank 1, and then passed through two combing reeds 10 and then guided by the two guide rollers 11 at the downstream end of the solution tank 1 to leave the solution tank 1, and enter a pair of rollers 6 above the waste liquid tank 5 for drying, and then the tension is adjusted by the downstream tensioner 9, and finally wound on the winding shaft 302. When the motor 303 is started and drives the winding shaft 302 to rotate, the winding shaft 302 pulls the sizing yarn from upstream to downstream.

[0060] After the sizing yarn 7 is immersed in the I2-KI solution, the sizing film on the sizing yarn 7 reacts with the I2-KI solution to develop color, the position of the high-speed camera 401 is adjusted to be located in the area after the sizing film color is stable, and the two strip light sources 402 are turned on for supplementary light, and then the sizing yarn 7 in the moving process is photographed to obtain 60 consecutive sizing yarn strip sequence images. The interval between each photo shooting by the high-speed camera 401 is not less than the time required for the sizing yarn in the camera shooting range to completely move out of the shooting range.

[0061] After the sizing yarn 7 is squeezed dry by the roller 6, the squeezed solution is collected in the waste liquid tank 5.

[0062] In order to avoid the influence of sunlight and other light on the shooting effect, the whole device can also be set in a black box to further improve the accuracy and stability of the photographed sizing yarn dryness sequence image and improve the calculation accuracy of the sizing yarn surface coating characteristic index.

[0063] The above embodiments are only illustrative of the principles and effects of the invention, as well as some embodiments of its application, and are not intended to limit the invention. It should be pointed out that a person skilled in the art can make several modifications and improvements without departing from the inventive concept of the invention, and all of these belong to the protection scope of the invention.

Claims

1. A method for calculating the surface coating characteristic index of sizing yarn based on image processing technology, characterized in that: The following steps are involved: (1) coloring the sizing film of the sizing yarn; (2) photographing the sizing yarn in sequence along the length direction to obtain a sequence of images of not less than 40 sizing yarns; (3) graying, noise reduction, background removal, and tilt correction are performed on the above-mentioned sizing yarn sequence images in turn by computer image processing technology to obtain a certain number of sizing yarn sequence grayscale images; (4) extracting an image of the effective statistical area of ​​the sizing yarn from each grayscale image of the sizing yarn sequence; (5) stitching a certain number of effective statistical area images of sizing yarn to form a long segment image of sizing yarn; (6) Processing the original yarn using the method described in steps (2) to (5) above to obtain a long segment image of the original yarn; (7) making a grayscale distribution histogram of all pixels in the long segment image of the sizing yarn and performing normal distribution fitting to obtain a normal distribution curve of grayscale values ​​of pixels in the sizing yarn; making a grayscale distribution histogram of all pixels in the long segment image of the original yarn and performing normal distribution fitting to obtain a normal distribution curve of grayscale values ​​of pixels in the original yarn; taking the intersection of the normal distribution curve of grayscale values ​​of pixels in the sizing yarn and the normal distribution curve of grayscale values ​​of pixels in the original yarn as the threshold T for segmenting the covered area, and the area in the grayscale distribution histogram of the sizing yarn where the grayscale value is lower than the threshold T is the area where the color reaction occurs, that is, the covered area; (8) Through the formula The grayscale distribution histogram of the effective statistical area of ​​the sizing yarn and the original yarn is converted from normal distribution to standard normal distribution; where Y represents the threshold in the standard normal distribution, T represents the normal distribution threshold, represents the average gray value of the pixels in the sizing long segment image, and σ represents the standard deviation of the gray value in the sizing long segment image; then, by looking up the "standard normal distribution table", the probability value less than the threshold value Y is obtained, that is, the sizing coverage coefficient SCC; (9) Through the formula Calculate the coating depth and use the coating depth SCD to evaluate the coating thickness of the sizing yarn; where G i Represents the gray value of the i-th pixel in the sizing long segment image, N y Indicates the number of pixels in the sizing long segment image; (10) Through the formula Calculate the coating unevenness and evaluate the coating unevenness of the sizing yarn with the coating unevenness SCU; where G i represents the gray value of the i-th pixel in the sizing long segment image, It represents the average gray value of pixels in the sizing long segment image; The specific method for extracting the effective statistical area image of the sizing yarn in step (4) is: respectively calculate the average grayscale value of each row of pixels in each grayscale image of the sizing yarn sequence, and select the area where the difference between the maximum average grayscale value and the minimum average grayscale value is less than 2 and the number of consecutive rows is greater than 10 as the effective statistical area image of the sizing yarn.

2. The method for calculating the surface coating characteristic index of sizing yarn based on image processing technology according to claim 1 is characterized in that: In step (3), the specific method of noise reduction is median filtering denoising; the specific method of background removal is: binarization is used to segment the image, the background other than the sizing yarn is removed, and the sizing yarn dryness image is obtained; the specific method of tilt correction is: the image is tilted and corrected by Hough transform.

3. The method for calculating the surface coating characteristic index of sizing yarn based on image processing technology according to claim 1, characterized in that: In the verification step (7), the grayscale distribution of all pixels in the sizing long segment image conforms to the normal distribution, and the grayscale distribution of all pixels in the original yarn long segment image conforms to the normal distribution.

4. A device for collecting a sequence of images of sizing yarn, wherein the sequence of images of sizing yarn is used in a method for calculating a characteristic index of sizing yarn surface coating based on image processing technology as claimed in claim 1, comprising a solution tank (1), wherein an unwinding device (2) and a winding device (3) are respectively arranged at two ends of the solution tank (1), a shooting device (4) is arranged above the solution tank (1), and the end of the solution tank (1) close to the unwinding device (2) is the upstream end, and the end close to the winding device (3) is the downstream end, characterized in that: A waste liquid tank (5) is also provided between the solution tank (1) and the winding device (3), and a pair of rollers (6) are provided on the waste liquid tank (5); The photographing device (4) comprises a high-speed camera (401) and two strip-shaped light sources (402) respectively arranged upstream and downstream of the high-speed camera (401); the high-speed camera (401) and the two strip-shaped light sources (402) are both mounted on a photographing frame (403), and the photographing frame (403) is fixedly arranged on one side of the solution tank (1); The unwinding device (2) comprises a plurality of unwinding tubes (201) arranged vertically and perpendicular to the moving direction of the sizing yarn (7); an L-shaped support plate (202) is arranged at one end of the arrangement direction of the unwinding tubes (201); the upper end of the support plate (202) is higher than the top of the unwinding tube (201) and is connected to a horizontally arranged first yarn guide roller (203); the axial direction of the first yarn guide roller (203) is parallel to the arrangement direction of each unwinding tube (201); the first yarn guide roller (203) is located directly above each unwinding tube (201); the lower part of the support plate (202) is connected to a first yarn guide roller (203) which is parallel to the first yarn guide roller (203); The winding device (3) comprises a second yarn guide roller (204) arranged between the unwinding tube (201) and the solution tank (1); the winding device (3) comprises shaft seats (301) arranged on both sides of the solution tank (1) and a winding shaft (302) rotatably connected between the two shaft seats (301), wherein a motor (303) is connected to the outer side of one of the shaft seats (301), and the motor (303) is transmission-connected to the winding shaft (302) to drive the winding shaft (302) to rotate; two combing reeds (10) are arranged in the solution tank (1), and the shooting device (4) is located between the two combing reeds (10).

5. The device for collecting sizing yarn sequence images according to claim 4, characterized in that: An upstream tensioner (8) for applying tension to the sizing yarn (7) is arranged between the unwinding device (2) and the solution tank (1), and a downstream tensioner (9) for applying tension to the sizing yarn (7) is arranged between the waste liquid tank (5) and the winding device (3).

Images