Method for checking quality standard of flat, multi-layer contoured textile fabric

Abstract

The invention provides a method for inspecting the quality standard of a planar textile tissue made of multiple layers of contours. The method comprises the following steps: acquiring a surface image of the textile tissue in real time through an image acquisition device; segmenting the textile tissue surface image through a multi-thread technology in combination with a preset window size to obtain a segmented image; on the basis of the segmented image, detecting whether the textile weave has defects in real time, if yes, recording a current textile weave surface image, and recording that the quality standard of the textile weave corresponding to the current textile weave surface image is unqualified; according to the recorded current textile tissue surface image, identifying defect types; and repairing the textile tissues with unqualified recorded quality standards. According to the invention, the efficiency and accuracy of the textile tissue quality standard inspection can be effectively improved.

Description

Technical Field

[0001] The present application relates to the field of industrial production technology, specifically to technical fields such as image processing, and can be applied in the context of textile manufacturing, and in particular to a method for inspecting the quality standards of a flat textile structure made of multi-layer contours. Background Art

[0004] At present, the quality standard inspection method of textile tissue is carried out manually, that is, it relies on human eyes to find textile tissue defects to further evaluate whether the textile tissue meets the corresponding quality standards. The human eye can accurately find the defect location of textiles, but due to the monotony and tediousness of the inspection work itself, and the human eye is easily fatigued, as the inspection time increases, the more tired the human eye is, the more mistakes will be made. According to statistics, even for skilled inspectors, the reliable efficiency of manual inspection of textile defects is difficult to reach 70%.

[0005] In addition, due to the limitations of the physiological functions of the human eye, the inspection work is slow and time-consuming, and it is difficult to meet the rapid production requirements of the modern textile industry. Summary of the invention

[0006] The present application provides a method for inspecting the quality standards of flat textile structures made with multi-layer profiles, which can reduce manual participation, effectively improve the efficiency and accuracy of textile structure quality standard inspection, and meet the rapid production requirements of modern textile industry.

[0007] According to a first aspect of the present application, a method for checking the quality standard of a flat textile structure made with a multi-layer profile is provided, the method comprising:

[0008] At a first preset position on the textile production line, a textile tissue surface image is acquired in real time by an image acquisition device;

[0009] Through multi-threading technology and preset window size, the textile tissue surface image is segmented to obtain the segmented image;

[0010] Based on the segmented image, a first preset algorithm is used to check in real time whether the textile tissue has defects, and when defects exist, a current textile tissue surface image is recorded, and a quality standard of the textile tissue corresponding to the current textile tissue surface image is recorded as unqualified;

[0011] When the textile tissue passes through the first preset position and arrives at the second preset position on the textile production line, the defect type is identified by a second preset algorithm according to the recorded current textile tissue surface image;

[0012] When the textile tissue passes through the second preset position and arrives at the third preset position on the textile production line, the repair strategy is started according to the third preset algorithm to repair the textile tissue that does not meet the recorded quality standard.

[0013] In some embodiments, the textile tissue surface image is segmented by multi-threading technology combined with a preset window size to obtain a segmented image, wherein the textile tissue surface image obtained on the textile production line is segmented by a window with a pixel size of 256×256.

[0014] In some embodiments, the method of inspecting in real time whether the textile tissue has defects based on the segmented image by a first preset algorithm, and when defects exist, recording the current textile tissue surface image, and recording that the quality standard of the textile tissue corresponding to the current textile tissue surface image is unqualified, wherein the method of inspecting in real time whether the textile tissue has defects based on the segmented image by a first preset algorithm specifically includes the following contents:

[0015] The segmented image is converted from a first color space to a second color space, and each channel of the second color space is subjected to homomorphic filtering to obtain a first processed image;

[0016] Performing roughness calculation on the first processed image, and dividing the first processed image into image blocks of the same size and non-overlapping according to the roughness calculation value;

[0017] The color distance difference between each image block and its adjacent image blocks is estimated respectively. When the difference is greater than a preset threshold, it is preliminarily determined that the corresponding image block has defects.

[0018] The image blocks initially judged to have defects are processed for saliency, and whether the corresponding image blocks have defects is judged again based on the processing results.

[0019] In some embodiments, the segmented image is converted from a first color space to a second color space, and each channel of the second color space is subjected to homomorphic filtering, wherein the first color space is an RGB color space and the second color space is an HSV color space.

[0020] In some embodiments, the roughness calculation is performed on the first processed image, and the first processed image is divided into image blocks of the same size and non-overlapping according to the roughness calculation value.

[0021] In some embodiments, when the textile tissue passes through the first preset position and arrives at the second preset position on the textile production line, the defect type is identified according to the recorded current textile tissue surface image by a second preset algorithm, wherein the defect type is identified according to the recorded current textile tissue surface image by a second preset algorithm, including the following contents:

[0022] Performing preset processing on the recorded current textile tissue surface image to obtain a purified image;

[0023] Perform morphological processing on the cleaned image to obtain the defect contour map;

[0024] The area of ​​the defect contour map is calculated, and the calculated area value is compared with the preset value to distinguish between hole-shaped defects and linear defects.

[0025] In some embodiments, the recorded current textile tissue surface image is subjected to a preset process to obtain a purified image, specifically:

[0026] The recorded current textile tissue surface image is processed by mean filtering, histogram averaging and variance threshold filtering to eliminate the interference of texture and noise points, and then obtain a purified image.

[0027] In some embodiments, the purified image is morphologically processed to obtain a defect contour map, wherein the morphological processing is to perform background correction on the purified image through mathematical morphology, which is used to extract image components that are meaningful for expressing and depicting the shape of the area from the image, so that subsequent recognition work can grasp the most essential morphological features of the target object.

[0028] In some embodiments, the area calculation of the defect contour map is performed, and the hole-shaped defect or the linear defect is distinguished by comparing the calculated area value with a preset value, which specifically includes the following contents:

[0029] Calculate the maximum connected domain area of ​​the defect contour map. When the maximum connected domain area is greater than the first preset threshold, it means that the current defect is a hole defect, determine its position and set the image number i`=i+1. Otherwise, calculate the horizontal and vertical variances and determine whether the variance is greater than the second preset threshold. When the calculation result is greater than the second preset threshold, it means that the current defect is a linear defect, determine its position and set the image number i`=i+1. Otherwise, it means that the textile structure here is normal, and set the image number i`=i+1.

[0030] Determine whether the image number satisfies i greater than n, where n is the pixel size value of the image. If it is greater than n, it means that the defect type has been identified and the process ends.

[0031] In some embodiments, when the textile tissue passes through the second preset position and arrives at the third preset position on the textile production line, the repair strategy is started according to the third preset algorithm to repair the textile tissue that does not meet the recorded quality standard, wherein the repair strategy is started according to the third preset algorithm to repair the textile tissue that does not meet the recorded quality standard, including the following contents:

[0032] Depending on the type of defect, different means are used to fit the missing part;

[0033] The image block at the edge of the defect is calculated by using the structural information of the current textile tissue surface image, and the textile tissue corresponding to the image block at the edge is used as the current block to be repaired;

[0034] Calculate multiple candidate matching blocks according to color, gradient and boundary features, and select the best matching block from them;

[0035] The current block to be repaired is filled by splitting and copying the best matching block, and it is iterated continuously until all missing parts are repaired.

[0036] Obviously, the embodiment of the present application obtains the surface image of the textile tissue in real time through the image acquisition device, and analyzes and processes the surface image of the textile tissue to check whether the textile tissue has defects, and then evaluates whether the quality standard of the textile tissue is qualified, thereby replacing the inspection method of the human eye. Therefore, it can reduce manual participation, effectively improve the efficiency and accuracy of the inspection of the quality standard of the textile tissue, and meet the rapid requirements of modern textile industry production.

[0037] At the same time, the embodiment of the present application presets different positions on the textile production line, which can not only adapt to the actual textile scene requirements, but also reserve sufficient time for the calculation of various algorithms, and reserve repair time and space for textile tissues with unqualified quality, thereby effectively improving the qualification rate of textile tissues.

[0038] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present application.

[0040] Figure 1 A schematic flow chart of a method for inspecting the quality standard of a flat textile structure made of a multi-layer profile provided in an embodiment of the present application. DETAILED DESCRIPTION

[0041] The following is a description of exemplary embodiments of the present application in conjunction with the accompanying drawings, including various details of the embodiments of the present application to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present application. Similarly, for the sake of clarity and conciseness, the description of well-known functions and structures is omitted in the following description.

[0042] It should be understood that in the embodiments of the present application, the character " / " generally indicates that the objects associated with each other are in an "or" relationship. The terms "first", "second", etc. are used only for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features.

[0043] The present application provides a method for inspecting the quality standards of flat textile structures made with multi-layer profiles, which can reduce manual participation, effectively improve the efficiency and accuracy of textile structure quality standard inspection, and meet the rapid production requirements of modern textile industry.

[0044] For example, the execution subject of the method for checking the quality standard of a flat textile structure made of a multi-layer profile can be a computer or a server, or can also be other equipment with data processing capabilities. The execution subject of the method is not limited here.

[0045] In some embodiments, the server may be a single server, or a server cluster consisting of multiple servers. In some implementations, the server cluster may also be a distributed cluster. The present application does not limit the specific implementation of the server.

[0046] Figure 1 A schematic flow chart of a method for testing the quality standard of a flat textile structure made of a multi-layer profile provided in an embodiment of the present application. Figure 1 As shown, the method may include the following steps:

[0047] Step 1, at a first preset position on a textile production line, a textile tissue surface image is acquired in real time by an image acquisition device;

[0048] Step 2, segmenting the textile tissue surface image by multi-threading technology combined with a preset window size to obtain a segmented image;

[0049] Step 3, based on the segmented image, using a first preset algorithm, real-time inspection of whether the textile tissue has defects, when defects exist, recording the current textile tissue surface image, and recording that the quality standard of the textile tissue corresponding to the current textile tissue surface image is unqualified;

[0050] Step 4, when the textile tissue passes through the first preset position and arrives at the second preset position on the textile production line, the defect type is identified according to the recorded current textile tissue surface image by using a second preset algorithm;

[0051] Step 5: When the textile tissue passes through the second preset position and arrives at the third preset position on the textile production line, a repair strategy is initiated according to a third preset algorithm to repair the textile tissue that does not meet the recorded quality standard.

[0052] Obviously, the embodiment of the present application obtains the surface image of the textile tissue in real time through the image acquisition device, and analyzes and processes the surface image of the textile tissue to check whether the textile tissue has defects, and then evaluates whether the quality standard of the textile tissue is qualified, thereby replacing the inspection method of the human eye. Therefore, it can reduce manual participation, effectively improve the efficiency and accuracy of the inspection of the quality standard of the textile tissue, and meet the rapid requirements of modern textile industry production.

[0053] At the same time, the embodiment of the present application presets different positions on the textile production line, which can not only adapt to the actual textile scene requirements, but also reserve sufficient time for the calculation of various algorithms, and reserve repair time and space for textile tissues with unqualified quality, thereby effectively improving the qualification rate of textile tissues.

[0054] In an embodiment of the present application, step 2, through multi-threading technology, combined with a preset window size, the textile tissue surface image is segmented to obtain a segmented image, wherein the textile tissue surface image obtained on the textile production line can be segmented through a window with a pixel size of 256×256.

[0055] Obviously, since the segmentation operation is performed through multi-threading technology, the processing speed and effect of the image can be effectively improved.

[0056] In the embodiment of the present application, step 3, based on the segmented image, by using a first preset algorithm, real-time inspection of whether the textile tissue has defects, when there are defects, recording the current textile tissue surface image, and recording that the quality standard of the textile tissue corresponding to the current textile tissue surface image is unqualified, wherein, based on the segmented image, by using a first preset algorithm, real-time inspection of whether the textile tissue has defects, specifically may include the following contents:

[0057] The segmented image is converted from a first color space to a second color space, and each channel of the second color space is subjected to homomorphic filtering to obtain a first processed image;

[0058] Performing roughness calculation on the first processed image, and dividing the first processed image into image blocks of the same size and non-overlapping according to the roughness calculation value;

[0059] The color distance difference between each image block and its adjacent image blocks is estimated respectively. When the difference is greater than a preset threshold, it is preliminarily determined that the corresponding image block has defects.

[0060] The image blocks initially judged to have defects are processed for saliency, and whether the corresponding image blocks have defects is judged again based on the processing results.

[0061] Obviously, the concept of image defect inspection that uses two different features to make preliminary judgments and re-verifications can achieve accurate defect inspection and avoid interference with the inspection results caused by the texture or contour of the textile structure itself.

[0062] For example, the segmented image is converted from a first color space to a second color space, and each channel of the second color space is subjected to homomorphic filtering, that is, through the conversion and filtering of the color space, when a defect exists, the contrast between the defect and the background can be improved, wherein the first color space can be an RGB color space, and the second color space can be an HSV color space.

[0063] For example, the roughness calculation is performed on the first processed image, and the first processed image is divided into image blocks of the same size and non-overlapping according to the roughness calculation value, wherein the roughness calculation can be expressed by the following formula:

[0064]

[0065] In the above formula, Q represents the roughness calculation value, m×n represents the pixel size of the image, and (x, y) is the pixel point.

[0066] For example, the color distance difference between each image block and its adjacent image block is estimated respectively. When the difference is greater than a preset threshold, it is preliminarily determined that the corresponding image block has defects. The color distance difference can be expressed by the following formula:

[0067]

[0068] In the above formula, N(D, O) represents the color distance difference between image blocks D and O, (O x , O y , O z ) represents the color value of image block O.

[0069] For example, the image blocks initially judged to have defects are processed for saliency, and the corresponding image blocks are judged again for defects based on the processing results. The saliency processing can be expressed by the following formula:

[0070] I s =||I μ -I w ||

[0071] In the above formula, I s Represents the eigenvalue of the image after saliency processing, I μ represents the mean image feature vector; I w It is represented by the pixel feature value of the image after Gaussian smoothing of the image block that is initially judged to have defects.

[0072] In addition, there are many ways to record the failure of the quality standard of the textile tissue corresponding to the current textile tissue surface image. For example, it can be recorded by the current size position of the textile tissue. Specifically, the whole textile tissue has a first section and a tail section, and the current position is 3 meters away from the first section. Of course, the textile tissue can also be divided into area numbers and the number of the current textile tissue can be recorded. The specific settings can be made according to needs and are not limited here.

[0073] In the embodiment of the present application, in step 4, when the textile tissue passes through the first preset position and arrives at the second preset position on the textile production line, the defect type is identified according to the recorded current textile tissue surface image by a second preset algorithm, wherein the defect type is identified according to the recorded current textile tissue surface image by a second preset algorithm, which may include the following steps:

[0074] Performing preset processing on the recorded current textile tissue surface image to obtain a purified image;

[0075] Perform morphological processing on the cleaned image to obtain the defect contour map;

[0076] The area of ​​the defect contour map is calculated, and the calculated area value is compared with the preset value to distinguish between hole-shaped defects and linear defects.

[0077] In practice, there are many defects in textile tissues, such as missing warp, broken warp, thick warp, reed marks, double warp, wrong threading, hanging warp, loose warp, loose weft, double weft, thick sections, thin sections, wrong weft, crotch change, holes and oil stains. However, they can be generally divided into two categories, namely: linear defects, such as missing warp, broken warp, thick warp, wrong weft, etc., and hole defects, such as holes, etc.

[0078] How to classify the defect types in actual application can be done according to needs and is not limited here.

[0079] Obviously, by performing morphological processing on the purified image, the defect contour map can be quickly obtained, and by calculating the area of ​​the defect contour map, the defect types can be quickly distinguished, thereby providing basic data for quality standard inspection and enabling targeted defect repair.

[0080] For example, the recorded current textile tissue surface image is processed by a preset process to obtain a purified image, which may be:

[0081] The recorded current textile tissue surface image is processed by mean filtering, histogram averaging and variance threshold filtering to eliminate the interference of texture and noise points, and then obtain a purified image.

[0082] For example, the purified image is morphologically processed to obtain a defect contour map, wherein the morphological processing is to perform background correction on the purified image through mathematical morphology, which is used to extract image components that are meaningful for expressing and depicting the shape of the area from the image, so that subsequent recognition work can grasp the most essential morphological features of the target object.

[0083] In this embodiment, morphological processing is performed on the cleaned image to obtain a defect contour map, which may include the following contents:

[0084] Select a 5×5 rectangular kernel structure element and perform an erosion operation on the purified image to remove discrete noise and shrink the image boundary;

[0085] A 7×7 rectangular kernel structure element is selected to perform an opening operation on the image to obtain the defect contour map.

[0086] For example, the area of ​​the defect contour map is calculated, and the hole-shaped defect or the line-shaped defect is distinguished by comparing the calculated area value with the preset value. Specifically, the following steps may be included:

[0087] Calculate the maximum connected domain area of ​​the defect contour map. When the maximum connected domain area is greater than the first preset threshold, it means that the current defect is a hole defect, determine its position and set the image number i`=i+1. Otherwise, calculate the horizontal and vertical variances and determine whether the variance is greater than the second preset threshold. When the calculation result is greater than the second preset threshold, it means that the current defect is a linear defect, determine its position and set the image number i`=i+1. Otherwise, it means that the textile structure here is normal, and set the image number i`=i+1.

[0088] Determine whether the image number satisfies i greater than n, where n is the pixel size value of the image. If it is greater than n, it means that the defect type has been identified and the process ends.

[0089] In the embodiment of the present application, in step 5, when the textile tissue passes through the second preset position and arrives at the third preset position on the textile production line, the repair strategy is started according to the third preset algorithm to repair the textile tissue that does not meet the recorded quality standard, wherein the repair strategy is started according to the third preset algorithm to repair the textile tissue that does not meet the recorded quality standard, which may include the following steps:

[0090] According to the different types of defects, different means are used to fit the missing parts to restore the complete structure of the textile tissue. For example, when it is a linear defect, a linear means can be used for fitting, and when it is a hole defect, a second-order Bezier curve means can be used for fitting;

[0091] The image block at the edge of the defect is calculated by using the structural information of the current textile tissue surface image, and the textile tissue corresponding to the image block at the edge is used as the current block to be repaired;

[0092] Calculate multiple candidate matching blocks based on color, gradient and boundary features, and select the best matching block from them to reduce the randomness in the selection process;

[0093] The current block to be repaired is filled by splitting and copying the best matching block to reduce the coverage of the non-missing part, and iterates continuously until all missing parts are repaired.

[0094] Obviously, by calculating multiple candidate matching blocks according to color, gradient and boundary features, and selecting the best matching block from them, and then repairing the defect according to the best matching block, a more natural repair can be achieved.

[0095] The above specific implementations do not constitute a limitation on the protection scope of this application. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of this application should be included in the protection scope of this application.

Claims

1. A method for checking the quality standard of a flat textile structure produced with a multi-layer profile, the method comprising: At a first preset position on the textile production line, a textile tissue surface image is acquired in real time by an image acquisition device; Through multi-threading technology and preset window size, the textile tissue surface image is segmented to obtain the segmented image; Based on the segmented image, a first preset algorithm is used to check in real time whether the textile tissue has defects, and when defects exist, a current textile tissue surface image is recorded, and a quality standard of the textile tissue corresponding to the current textile tissue surface image is recorded as unqualified; When the textile tissue passes through the first preset position and arrives at the second preset position on the textile production line, the defect type is identified according to the recorded current textile tissue surface image by using a second preset algorithm; When the textile tissue passes through the second preset position and arrives at the third preset position on the textile production line, the repair strategy is started according to the third preset algorithm to repair the textile tissue that does not meet the recorded quality standard.

2. The method according to claim 1, wherein the textile tissue surface image is segmented by multi-threading technology combined with a preset window size to obtain a segmented image, wherein: The surface image of the textile structure obtained on the textile production line is segmented through a window with a pixel size of 256×256.

3. The method according to claim 1, wherein the segmented image is used to detect in real time whether the textile tissue has defects through a first preset algorithm, and when defects exist, the current textile tissue surface image is recorded, and the quality standard of the textile tissue corresponding to the current textile tissue surface image is recorded as unqualified, wherein: Based on the segmented image, a first preset algorithm is used to check in real time whether the textile structure has defects, which specifically includes the following contents: The segmented image is converted from a first color space to a second color space, and each channel of the second color space is subjected to homomorphic filtering to obtain a first processed image; Performing roughness calculation on the first processed image, and dividing the first processed image into image blocks of the same size and non-overlapping according to the roughness calculation value; The color distance difference between each image block and its adjacent image blocks is estimated respectively. When the difference is greater than a preset threshold, it is preliminarily determined that the corresponding image block has defects. The image blocks initially judged to have defects are processed for saliency, and whether the corresponding image blocks have defects is judged again based on the processing results.

4. The method according to claim 3, wherein the segmented image is converted from the first color space to the second color space, and each channel of the second color space is subjected to homomorphic filtering, wherein: The first color space is the RGB color space, and the second color space is the HSV color space.

5. The method according to claim 3, wherein the roughness calculation is performed on the first processed image, and the first processed image is divided into image blocks of the same size and non-overlapping according to the roughness calculation value.

6. The method according to claim 1, wherein when the textile tissue passes through the first preset position and arrives at the second preset position on the textile production line, the defect type is identified according to the recorded current textile tissue surface image by using a second preset algorithm, wherein: According to the recorded current textile structure surface image, the second preset algorithm is used to identify the defect types, including the following: Performing preset processing on the recorded current textile tissue surface image to obtain a purified image; Perform morphological processing on the cleaned image to obtain the defect contour map; The area of ​​the defect contour map is calculated, and the calculated area value is compared with the preset value to distinguish between hole-shaped defects and linear defects.

7. The method according to claim 6, wherein the recorded current textile tissue surface image is subjected to a preset process to obtain a purified image, specifically: The recorded current textile tissue surface image is processed by mean filtering, histogram averaging and variance threshold filtering to eliminate the interference of texture and noise points, and then obtain a purified image.

8. The method according to claim 6, wherein the cleaned image is subjected to morphological processing to obtain a defect contour map, wherein: Morphological processing is to perform background correction on the purified image through mathematical morphology, which is used to extract image components that are meaningful for expressing and depicting the shape of the region, so that subsequent recognition work can grasp the most essential morphological features of the target object.

9. The method according to claim 6, wherein the area of ​​the defect contour map is calculated, and the hole defect or the linear defect is distinguished by comparing the calculated area value with a preset value, which specifically includes the following contents: Calculate the maximum connected domain area of ​​the defect contour map. When the maximum connected domain area is greater than the first preset threshold, it means that the current defect is a hole defect, determine its position and set the image number i`=i+1. Otherwise, calculate the horizontal and vertical variances and determine whether the variance is greater than the second preset threshold. When the calculation result is greater than the second preset threshold, it means that the current defect is a linear defect, determine its position and set the image number i`=i+1. Otherwise, it means that the textile structure here is normal, and set the image number i`=i+1. Determine whether the image number satisfies i greater than n, where: n is the pixel size of the image. When it is greater than n, it means that the defect type has been identified and the process ends.

10. The method according to claim 1, wherein when the textile tissue passes through the second preset position and arrives at the third preset position on the textile production line, a repair strategy is initiated according to a third preset algorithm to repair the textile tissue that does not meet the recorded quality standard, wherein: The repair strategy is initiated according to the third preset algorithm to repair the textile tissue that does not meet the recorded quality standards, including the following contents: Depending on the type of defect, different means are used to fit the missing part; The image block at the edge of the defect is calculated by using the structural information of the current textile tissue surface image, and the textile tissue corresponding to the image block at the edge is used as the current block to be repaired; Calculate multiple candidate matching blocks according to color, gradient and boundary features, and select the best matching block from them; The current block to be repaired is filled by splitting and copying the best matching block, and it is iterated continuously until all missing parts are repaired.

Images