CN114926436A - Defect detection method for periodic pattern fabric - Google Patents
Defect detection method for periodic pattern fabric Download PDFInfo
- Publication number
- CN114926436A CN114926436A CN202210553548.3A CN202210553548A CN114926436A CN 114926436 A CN114926436 A CN 114926436A CN 202210553548 A CN202210553548 A CN 202210553548A CN 114926436 A CN114926436 A CN 114926436A
- Authority
- CN
- China
- Prior art keywords
- fabric
- image
- gray scale
- gray
- global
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 239000004744 fabric Substances 0.000 title claims abstract description 131
- 230000007547 defect Effects 0.000 title claims abstract description 40
- 238000001514 detection method Methods 0.000 title claims abstract description 20
- 230000000737 periodic effect Effects 0.000 title claims abstract description 11
- 238000000034 method Methods 0.000 claims abstract description 42
- 230000008859 change Effects 0.000 claims abstract description 35
- 238000001228 spectrum Methods 0.000 claims abstract description 30
- 230000002159 abnormal effect Effects 0.000 claims abstract description 27
- 230000011218 segmentation Effects 0.000 claims description 22
- 238000000638 solvent extraction Methods 0.000 claims 1
- 238000013473 artificial intelligence Methods 0.000 abstract description 2
- 238000004364 calculation method Methods 0.000 description 6
- 230000005856 abnormality Effects 0.000 description 4
- 238000010606 normalization Methods 0.000 description 4
- 238000005070 sampling Methods 0.000 description 3
- GVJHHUAWPYXKBD-UHFFFAOYSA-N (±)-α-Tocopherol Chemical compound OC1=C(C)C(C)=C2OC(CCCC(C)CCCC(C)CCCC(C)C)(C)CCC2=C1C GVJHHUAWPYXKBD-UHFFFAOYSA-N 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 239000000843 powder Substances 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 229930003427 Vitamin E Natural products 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- WIGCFUFOHFEKBI-UHFFFAOYSA-N gamma-tocopherol Natural products CC(C)CCCC(C)CCCC(C)CCCC1CCC2C(C)C(O)C(C)C(C)C2O1 WIGCFUFOHFEKBI-UHFFFAOYSA-N 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
- 229940046009 vitamin E Drugs 0.000 description 1
- 235000019165 vitamin E Nutrition 0.000 description 1
- 239000011709 vitamin E Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
- G06T7/0008—Industrial image inspection checking presence/absence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/40—Analysis of texture
- G06T7/41—Analysis of texture based on statistical description of texture
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30124—Fabrics; Textile; Paper
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Probability & Statistics with Applications (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to the technical field of artificial intelligence, in particular to a defect detection method for a periodic pattern fabric. The method comprises the steps of obtaining a frequency spectrum image of each frame of fabric image; acquiring global gray scale characteristics of the frequency spectrum image according to local gray scale change information of each pixel point on the diagonal of the frequency spectrum image; based on the periodicity and consistency of fabric patterns, calculating the gray information abnormal degree of each frame of fabric image according to the image acquisition time and the global gray characteristic; and confirming the fabric defect type by the global gray feature based on the abnormal degree of the gray information. The defect detection is carried out based on the frequency spectrum image, so that the difference of the periodicity of the fabric patterns on the airspace can be avoided, the defect condition of the fabric can be detected by utilizing the periodicity and the consistency of the fabric patterns, the limitation of background modeling can be solved, and the defect detection efficiency and the accuracy of the detection result are improved.
Description
Technical Field
The invention relates to the technical field of artificial intelligence, in particular to a defect detection method for a periodic pattern fabric.
Background
Most of the patterns on the fabric are printed and dyed by using a specific mold, so the patterns on the fabric are in a periodic reciprocating form. For the integrity and aesthetic appearance of the fabric pattern, defect detection for the fabric pattern is increasingly important.
At present, most of fabric pattern defect detection methods utilize an image processing mode, namely, a background modeling mode to detect pattern defects in real time, but the spatial domain characteristics of acquired fabric images are different, and the produced cloth patterns are variable, so that the fabric in each pattern is difficult to perform corresponding background modeling, therefore, the background modeling method has limitations and can cause the reduction of detection efficiency.
Disclosure of Invention
In order to solve the above technical problems, an object of the present invention is to provide a method for detecting defects of a periodically patterned fabric, which adopts the following technical solutions:
acquiring multiple frames of fabric images based on a time sequence, and converting each frame of fabric image into a gray image to obtain a frequency spectrum image of the gray image;
simplifying the frequency spectrum image into a binary image by using a maximum inter-class threshold method, acquiring any one half diagonal line of diagonal lines of the binary image by taking a central point of the binary image as an origin, respectively taking each pixel point on the half diagonal line as an angular point of a corresponding rectangular region, respectively acquiring local gray scale change information corresponding to each pixel point on the half diagonal line according to pixel values of edge pixel points of the rectangular region, and acquiring global gray scale characteristics of the binary image according to the local gray scale change information;
based on the periodicity and consistency of fabric patterns, acquiring the irregularity degree of the patterns of each frame of fabric images by combining the acquisition time corresponding to all the fabric images, and calculating the outlier degree of each frame of fabric images corresponding to the global gray-scale features by using the global gray-scale features corresponding to all the fabric images; combining the irregularity degree of the patterns and the outlier degree to obtain the abnormal degree of the gray information corresponding to the fabric image;
and confirming the fabric defect type by the global gray feature based on the abnormal degree of the gray information.
Further, the method for acquiring the binary image comprises the following steps:
and acquiring the optimal segmentation threshold value of the frequency spectrum image by using a maximum inter-class threshold value method, setting the pixel value which is greater than or equal to the optimal segmentation threshold value in the frequency spectrum image as 1, and setting the pixel value which is less than the optimal segmentation threshold value in the frequency spectrum image as 0.
Further, the method for respectively obtaining local gray scale change information corresponding to each pixel point on the half diagonal line according to the pixel value of the edge pixel point of the rectangular region includes:
counting the number of the pixel values of the edge pixel points in the rectangular area corresponding to the pixel points on the half diagonal line as 1 and the total number of the edge pixel points in the rectangular area; and calculating the ratio of the number to the total number, and taking the ratio as the local gray scale change information of the corresponding pixel points on the half diagonal.
Further, the method for acquiring the global gray scale feature includes:
forming the local gray scale change information of each pixel point on the half diagonal into a diagonal description sequence, and acquiring segmentation points of a high-frequency position and a low-frequency position according to the difference between the local gray scale change information in the diagonal description sequence;
and obtaining high-frequency information and low-frequency information in the binary image according to the segmentation point, calculating a ratio between the high-frequency information and the low-frequency information, and taking the ratio as the global gray feature corresponding to the fabric image.
Further, the method for obtaining the segmentation point includes:
obtaining a difference value sequence corresponding to the diagonal description sequence, wherein each difference value in the difference value sequence is obtained by a difference value between adjacent local gray scale change information in the diagonal description sequence;
and taking the previous local gray scale change information in the two local gray scale change information corresponding to the maximum difference value in the difference value sequence as a segmentation point.
Further, the method for obtaining high frequency information and low frequency information according to the dividing point includes:
and based on the diagonal description sequence, taking the sum of all local gray scale change information on the left side of the division point as the low-frequency information and the sum of all local gray scale change information on the right side of the division point as the high-frequency information.
Further, the method for acquiring the irregularity degree of the patterns of each frame of the fabric images by combining the acquisition time corresponding to all the fabric images comprises the following steps:
acquiring the global gray feature of each frame of fabric image, and forming a global gray feature sequence by the acquired multiple frames of fabric images according to the acquisition time sequence of the fabric images;
performing one-dimensional density clustering on the global gray features in the global gray feature sequence to obtain a plurality of clustered categories; acquiring the average time interval of all the fabric images in the same category according to the acquisition time corresponding to each frame of fabric image;
and calculating the irregularity degree of the pattern of the current frame by combining the time difference of corresponding acquisition time of the current frame and the left and right adjacent frames in the same category of the current frame and the average time interval of the category of the current frame.
Further, the method for calculating the degree of outlier of the fabric image corresponding to the global gray feature in each frame by using the global gray features corresponding to all the fabric images includes:
acquiring the global gray scale features corresponding to the mode in the category to which the current frame belongs, and calculating the average difference between each global gray scale feature in the category and the global gray scale feature;
calculating the outlier degree of the current frame corresponding to the global gray feature from the average difference, the global gray feature corresponding to the mode, and the global gray feature of the current frame.
Further, the method for acquiring the abnormal degree of the gray scale information corresponding to the fabric image by combining the irregularity degree of the pattern and the outlier degree comprises the following steps:
and the product of the irregularity degree of the patterns and the outlier degree is the abnormal degree of the gray information corresponding to the fabric image.
Further, the method for confirming the fabric defect type comprises the following steps:
and setting an abnormal threshold, and when the abnormal degree of the gray information is greater than the abnormal threshold, determining that the fabric image has defects, and further judging the fabric defect type according to the global gray characteristic of the fabric image.
The embodiment of the invention at least has the following beneficial effects: the defect detection is carried out based on the frequency spectrum image, so that the difference of the periodicity of the fabric patterns on the airspace can be avoided, the defect condition of the fabric can be detected by utilizing the periodicity and the consistency of the fabric patterns, the limitation of background modeling can be solved, and the defect detection efficiency and the accuracy of the detection result are improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions and advantages of the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
FIG. 1 is a flowchart illustrating the steps of a method for detecting defects in a periodically patterned fabric, according to one embodiment of the present invention;
FIG. 2 is a schematic view of a scene of a camera capturing a fabric image according to an embodiment of the present invention;
fig. 3 is a schematic diagram of a rectangular area corresponding to a half diagonal pixel point according to an embodiment of the invention.
Detailed Description
To further illustrate the technical means and effects of the present invention adopted to achieve the predetermined objects, the following detailed description of the defect detection method for periodic patterned fabric according to the present invention with reference to the accompanying drawings and preferred embodiments, the detailed description, the structure, the features and the effects thereof, is provided below. In the following description, different "one embodiment" or "another embodiment" refers to not necessarily the same embodiment. Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The following describes a specific scheme of the defect detection method for periodic patterned fabric provided by the invention in detail with reference to the accompanying drawings.
The embodiment of the invention aims at the following specific scenes: the pattern of the fabric is periodic, but the period size is not fixed; the view angle and the focal length of the camera are fixed, and the size of the acquired image range is fixed.
Referring to fig. 1, a flow chart of the steps of a method for detecting defects in a periodically patterned fabric according to an embodiment of the present invention is shown, the method comprising the steps of:
and S001, collecting multiple frames of fabric images based on the time sequence, and converting each frame of fabric image into a gray level image to obtain a frequency spectrum image of the gray level image.
Specifically, referring to fig. 2, when a corresponding camera a is arranged on the fabric production line to collect fabric images, and the camera collecting range B is a fixed value, and the fabric pattern period C is random, multiple frames of fabric images are collected according to a time sequence.
The uncertainty of the fabric pattern period enables the collected fabric images to have different distribution relations on a space domain, but after the fabric images are converted into a frequency domain, the data have consistency, namely when the periodic content of the images is not changed, and only the positions are translated, the frequency spectrums of the images are not changed, therefore, the frequency spectrum images of each frame of fabric images are obtained, and the obtaining method comprises the following steps: and converting the fabric image into a gray image, performing two-dimensional discrete Fourier transform on the gray image, performing logarithm treatment and center treatment on a transform result to obtain a magnitude spectrum, namely a frequency spectrum image.
Step S002, the frequency spectrum image is simplified into a binary image by utilizing a maximum inter-class threshold method, any half diagonal line of the diagonal lines of the binary image is obtained by taking the central point of the binary image as the original point, each pixel point on the half diagonal line is respectively taken as the corner point of the corresponding rectangular area, the local gray scale change information corresponding to each pixel point on the half diagonal line is respectively obtained according to the pixel value of the edge pixel point of the rectangular area, and the global gray scale feature of the binary image is obtained according to the local gray scale change information.
Specifically, for fabric images with different pattern styles, corresponding grayscale images and spectrum images are different, in order to adopt a unified processing mode, normalization processing needs to be performed on the processed images, and even if pixel values in the spectrum images are all in the range of [0,1], the normalization formula is as follows:
wherein, g (u,v) The pixel values of the v-column pixel points in the u-th row after normalization processing; f. of (u,v) The pixel value of the pixel point of the v column of the u row before normalization processing; max (f) is the largest pixel value in the spectral image; min (f) is the minimum pixel value in the spectrum image.
In order to make subsequent calculation and description simpler and more efficient, the floating point type of data needs to be discarded, and the calculation and description are performed by adopting simple binary data as much as possible, then the maximum inter-class threshold method is used to obtain the optimal segmentation threshold value for performing the second classification on the normalized frequency spectrum image, the pixel value greater than or equal to the optimal segmentation threshold value is set as 1, and the pixel value smaller than the optimal segmentation threshold value is set as 0, so that the simplified binary image of the normalized frequency spectrum image can be obtained.
Further, after the binary image is obtained, information description needs to be performed on the image, so that description differences are compared subsequently to determine an abnormality, and then the information description method in the binary image is as follows: referring to FIG. 3, first, the diagonal length of the binary image is obtained
Wherein C represents the width of the binary image, H represents the height of the binary image; then, any half diagonal line in any diagonal line is obtained by taking the central point of the binary image as the origin, and each pixel point on the half diagonal line is taken as the originThe corner points of the rectangular area correspond to the local gray scale change information of the corresponding pixel points on the half diagonal line is calculated according to the pixel values of the edge pixel points of the rectangular area, and then the calculation formula is as follows:
wherein Ms n The local gray scale change information corresponding to the nth pixel point on the half diagonal line; SL (Long-side) n The number of pixel values of the edge pixel points in the rectangular area corresponding to the nth pixel point is 1; and N is the total number of edge pixel points in the rectangular area.
According to the length of the half diagonal line, local gray change information corresponding to each pixel point can be obtained to form a diagonal description sequence [ Ms 1 ,Ms 2 ,Ms 3 ,..,Ms M-1 ,Ms M ]And M is the total number of pixel points on the half diagonal.
Because the diagonal description sequence can reflect the gray information of high and low frequency positions with different degrees in the binary image, the diagonal description sequence is integrated to obtain the global gray feature in the binary image so as to quickly judge the abnormality from a higher macroscopic scale, and the acquisition process of the global gray feature is as follows:
(1) obtaining segmentation points of a high-frequency position and a low-frequency position according to the difference between local gray scale change information in the diagonal description sequence, and then firstly obtaining a difference sequence corresponding to the diagonal description sequence, wherein each difference in the difference sequence is obtained by the difference between adjacent local gray scale change information in the diagonal description sequence; and taking the previous local gray scale change information in the two local gray scale change information corresponding to the maximum difference value in the difference value sequence as a segmentation point.
(2) Taking the sum of all local gray scale change information on the left side of the division point as low-frequency information and the sum of all local gray scale change information on the right side of the division point as high-frequency information, and calculating the ratio of the high-frequency information to the low-frequency information
Wherein x represents the position of the segmentation point, and the ratio is used as the global gray feature of the corresponding fabric image.
Step S003, based on the periodicity and consistency of the fabric patterns, combining the acquisition time corresponding to all the fabric images to obtain the pattern irregularity degree of each frame of fabric image, and calculating the outlier degree of the global gray feature corresponding to each frame of fabric image by using the global gray features corresponding to all the fabric images; and obtaining the abnormal degree of the gray information of the corresponding fabric image by combining the irregularity degree and the outlier degree of the patterns.
Specifically, the method in steps S001 and S002 is used to obtain the global gray feature of each frame of fabric image, the collected multi-frame fabric images form a global gray feature sequence according to the collection time sequence, and the abnormality of the corresponding position of the fabric is determined according to the abnormality degree of the global gray feature in the global gray feature sequence, and then the method is as follows:
(1) and performing one-dimensional density clustering on the global gray features in the global gray feature sequence to obtain a plurality of clustered categories, and acquiring the average time interval of all fabric images in the same category according to the acquisition time corresponding to each frame of fabric image, wherein each category has an average time interval.
(2) Respectively calculating the irregularity degree of the patterns of each frame of fabric image according to the average time interval of each category, wherein the calculation formula is as follows:
wherein, gl t The irregularity degree of the pattern of the t frame fabric image; t is t Sampling time of the t frame fabric image; t is t-1 Sampling time of the t-1 th frame of fabric image in the same category; t is t+1 Sampling time of a t +1 th frame of fabric image in the same category;
representing the average time interval of the corresponding category of the t frame of fabric image; % is the operation of taking the rest.
(3) Calculating the outlier degree of the global gray feature corresponding to each frame of fabric image according to the global gray feature in the same category, wherein the calculation formula is as follows:
wherein lq is t The outlier degree of the global gray scale characteristic corresponding to the t frame of fabric image; pu (Pu) powder t The global gray scale characteristic of the t frame fabric image is obtained; pu (Pu) powder 0 Global gray scale features corresponding to the mode in the category to which the t-th frame of fabric image belongs;
the average difference between each global gray feature in the category to which the t frame fabric image belongs and the global gray feature corresponding to the mode is obtained; r is the number of global gray features in the category; pu (vitamin E) r And obtaining the global gray feature corresponding to the r-th frame of the fabric image.
(4) Combining the irregularity degree and the outlier degree to obtain the abnormal degree of the gray information of the corresponding fabric image, wherein the more irregular the abnormal degree is, the more outlier the abnormal degree is, the calculation formula of the abnormal degree of the gray information is as follows: yc t ＝gl t *lq t Wherein, yc t The abnormal degree of the gray information of the t frame fabric image is shown.
And step S004, confirming the fabric defect type by the global gray characteristic based on the abnormal degree of the gray information.
Specifically, an abnormal threshold value is set, when the abnormal degree of the gray information is greater than the abnormal threshold value, the defect of the corresponding fabric image is determined, and then the fabric defect type is judged according to the global gray feature of the fabric image: when the global gray feature is greater than or equal to the feature threshold, it is indicated that high-frequency information is increased, and the fabric is confirmed to have high-frequency defects such as warp and weft defects, defects and the like; and when the global gray scale feature is smaller than the feature threshold, indicating that the low-frequency information is increased, and confirming that the new defect of the fabric occurs in the area.
Preferably, the anomaly threshold value in the embodiment of the invention is 0.3.
In summary, the embodiment of the present invention provides a defect detection method for a periodic patterned fabric, which obtains a frequency spectrum image of each frame of fabric image; acquiring global gray scale characteristics of the frequency spectrum image according to local gray scale change information of each pixel point on the diagonal of the frequency spectrum image; based on the periodicity and consistency of fabric patterns, calculating the gray information abnormal degree of each frame of fabric image according to the image acquisition time and the global gray feature; and confirming the fabric defect type by the global gray feature based on the abnormal degree of the gray information. The defect detection is carried out based on the frequency spectrum image, so that the difference of the periodicity of the fabric patterns on the airspace can be avoided, the defect condition of the fabric can be detected by utilizing the periodicity and the consistency of the fabric patterns, the limitation of background modeling can be solved, and the defect detection efficiency and the accuracy of the detection result are improved.
It should be noted that: the precedence order of the above embodiments of the present invention is only for description, and does not represent the merits of the embodiments. And specific embodiments thereof have been described above. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.
All the embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from other embodiments.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and should not be taken as limiting the scope of the present invention, which is intended to cover any modifications, equivalents, improvements, etc. within the spirit and scope of the present invention.
Claims (10)
1. A defect detection method based on a periodic pattern fabric is characterized by comprising the following steps:
acquiring multiple frames of fabric images based on a time sequence, and converting each frame of fabric image into a gray image to obtain a frequency spectrum image of the gray image;
simplifying the frequency spectrum image into a binary image by using a maximum inter-class threshold method, acquiring any half diagonal line of the diagonal lines of the binary image by taking a central point of the binary image as an origin, respectively taking each pixel point on the half diagonal line as an angular point of a corresponding rectangular area, respectively acquiring local gray scale change information corresponding to each pixel point on the half diagonal line according to the pixel value of an edge pixel point of the rectangular area, and acquiring the global gray scale feature of the binary image according to the local gray scale change information;
based on the periodicity and consistency of fabric patterns, acquiring the pattern irregularity degree of each frame of fabric image by combining the acquisition time corresponding to all the fabric images, and calculating the outlier degree of each frame of fabric image corresponding to the global gray feature by using the global gray features corresponding to all the fabric images; combining the irregularity degree of the patterns and the outlier degree to obtain the abnormal degree of the gray information corresponding to the fabric image;
and confirming the fabric defect type by the global gray feature based on the abnormal degree of the gray information.
2. The method as claimed in claim 1, wherein the binary image obtaining method comprises:
and acquiring the optimal segmentation threshold value of the frequency spectrum image by using a maximum inter-class threshold value method, setting the pixel value which is greater than or equal to the optimal segmentation threshold value in the frequency spectrum image as 1, and setting the pixel value which is less than the optimal segmentation threshold value in the frequency spectrum image as 0.
3. The method according to claim 2, wherein the method for respectively obtaining local gray scale change information corresponding to each pixel point on the half diagonal line according to the pixel values of the edge pixel points of the rectangular region comprises:
counting the number of the edge pixel points in the rectangular region corresponding to the pixel points on the half diagonal line, wherein the pixel value of the edge pixel points is 1, and the total number of the edge pixel points in the rectangular region; and calculating the ratio between the number and the total number, and taking the ratio as the local gray scale change information of the corresponding pixel points on the half diagonal.
4. The method of claim 1, wherein the global gray scale feature obtaining method comprises:
forming the local gray scale change information of each pixel point on the half diagonal into a diagonal description sequence, and acquiring segmentation points of a high-frequency position and a low-frequency position according to the difference between the local gray scale change information in the diagonal description sequence;
and obtaining high-frequency information and low-frequency information in the binary image according to the segmentation points, calculating a ratio between the high-frequency information and the low-frequency information, and taking the ratio as the global gray feature corresponding to the fabric image.
5. The method of claim 4, wherein the method for obtaining the segmentation point comprises:
obtaining a difference value sequence corresponding to the diagonal description sequence, wherein each difference value in the difference value sequence is obtained by a difference value between adjacent local gray scale change information in the diagonal description sequence;
and taking the previous local gray scale change information in the two local gray scale change information corresponding to the maximum difference value in the difference value sequence as a segmentation point.
6. The method of claim 4, wherein the method of deriving high frequency information and low frequency information based on the partitioning point comprises:
and based on the diagonal description sequence, taking the sum of all local gray scale change information on the left side of the segmentation point as the low-frequency information and the sum of all local gray scale change information on the right side of the segmentation point as the high-frequency information.
7. The method of claim 1, wherein said step of obtaining a pattern irregularity for each frame of said fabric images in combination with corresponding acquisition times for all of said fabric images comprises:
acquiring the global gray scale feature of each frame of fabric image, and forming a global gray scale feature sequence by the acquired multiple frames of fabric images according to the acquisition time sequence of the fabric images;
performing one-dimensional density clustering on the global gray features in the global gray feature sequence to obtain a plurality of clustered categories; acquiring the average time interval of all the fabric images in the same category according to the corresponding acquisition time of each frame of fabric image;
and calculating the irregularity degree of the pattern of the current frame by combining the time difference of corresponding acquisition time of the current frame and the left and right adjacent frames in the same category of the current frame and the average time interval of the category of the current frame.
8. The method of claim 1, wherein the calculating the degree of outlier of the fabric image corresponding to the global gray feature for each frame using the global gray features corresponding to all the fabric images comprises:
acquiring the global gray scale features corresponding to the mode in the category to which the current frame belongs, and calculating the average difference between each global gray scale feature in the category and the global gray scale feature;
calculating the outlier of a current frame corresponding to the global gray feature from the average difference, the global gray feature corresponding to the mode, and the global gray feature of the current frame.
9. The method according to claim 1, wherein said combining said degree of pattern irregularity and said degree of outlier to obtain a degree of anomaly of gray scale information corresponding to said image of said web comprises:
and the product of the irregularity degree of the patterns and the outlier degree is the abnormal degree of the gray information corresponding to the fabric image.
10. The method according to claim 1, wherein the method of identifying the type of fabric defect comprises:
and setting an abnormal threshold, and when the abnormal degree of the gray information is greater than the abnormal threshold, determining that the fabric image has defects, and further judging the fabric defect type according to the global gray characteristic of the fabric image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210553548.3A CN114926436A (en) | 2022-05-20 | 2022-05-20 | Defect detection method for periodic pattern fabric |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210553548.3A CN114926436A (en) | 2022-05-20 | 2022-05-20 | Defect detection method for periodic pattern fabric |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114926436A true CN114926436A (en) | 2022-08-19 |
Family
ID=82811097
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210553548.3A Pending CN114926436A (en) | 2022-05-20 | 2022-05-20 | Defect detection method for periodic pattern fabric |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114926436A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115082467A (en) * | 2022-08-22 | 2022-09-20 | 山东亿昌装配式建筑科技有限公司 | Building material welding surface defect detection method based on computer vision |
| CN115239711A (en) * | 2022-09-21 | 2022-10-25 | 苏州琼派瑞特科技股份有限公司 | Online operation abnormity identification system of sewing equipment |
| CN115272323A (en) * | 2022-09-28 | 2022-11-01 | 南通羿云智联信息科技有限公司 | Data intelligent regulation and control acquisition method for traffic engineering pavement quality detection |
| CN115311278A (en) * | 2022-10-11 | 2022-11-08 | 南通欧惠纺织科技有限公司 | Yarn cutting method for yarn detection |
| CN115311267A (en) * | 2022-10-10 | 2022-11-08 | 南通逸耀辰纺织品科技有限公司 | Method for detecting abnormity of check fabric |
| CN115423807A (en) * | 2022-11-04 | 2022-12-02 | 山东益民服饰有限公司 | Cloth defect detection method based on outlier detection |
| CN115618051A (en) * | 2022-12-20 | 2023-01-17 | 楠楠聚智信息科技有限责任公司 | Internet-based smart campus monitoring video storage method |
| CN116805312A (en) * | 2023-08-21 | 2023-09-26 | 青岛时佳汇服装有限公司 | Knitted fabric quality detection method based on image processing |
| CN117152444A (en) * | 2023-10-30 | 2023-12-01 | 山东泰普锂业科技有限公司 | Equipment data acquisition method and system for lithium battery industry |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2019104767A1 (en) * | 2017-11-28 | 2019-06-06 | 河海大学常州校区 | Fabric defect detection method based on deep convolutional neural network and visual saliency |
| CN111179225A (en) * | 2019-12-14 | 2020-05-19 | 西安交通大学 | Test paper surface texture defect detection method based on gray gradient clustering |
| CN113643294A (en) * | 2021-10-14 | 2021-11-12 | 江苏祥顺布业有限公司 | Textile defect self-adaptive detection method based on frequency spectrum analysis |
-
2022
- 2022-05-20 CN CN202210553548.3A patent/CN114926436A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2019104767A1 (en) * | 2017-11-28 | 2019-06-06 | 河海大学常州校区 | Fabric defect detection method based on deep convolutional neural network and visual saliency |
| CN111179225A (en) * | 2019-12-14 | 2020-05-19 | 西安交通大学 | Test paper surface texture defect detection method based on gray gradient clustering |
| CN113643294A (en) * | 2021-10-14 | 2021-11-12 | 江苏祥顺布业有限公司 | Textile defect self-adaptive detection method based on frequency spectrum analysis |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115082467A (en) * | 2022-08-22 | 2022-09-20 | 山东亿昌装配式建筑科技有限公司 | Building material welding surface defect detection method based on computer vision |
| CN115239711B (en) * | 2022-09-21 | 2023-04-07 | 苏州琼派瑞特科技股份有限公司 | Online operation abnormity identification system of sewing equipment |
| CN115239711A (en) * | 2022-09-21 | 2022-10-25 | 苏州琼派瑞特科技股份有限公司 | Online operation abnormity identification system of sewing equipment |
| CN115272323A (en) * | 2022-09-28 | 2022-11-01 | 南通羿云智联信息科技有限公司 | Data intelligent regulation and control acquisition method for traffic engineering pavement quality detection |
| CN115272323B (en) * | 2022-09-28 | 2022-12-27 | 南通羿云智联信息科技有限公司 | Data intelligent regulation and control acquisition method for traffic engineering pavement quality detection |
| CN115311267A (en) * | 2022-10-10 | 2022-11-08 | 南通逸耀辰纺织品科技有限公司 | Method for detecting abnormity of check fabric |
| CN115311278A (en) * | 2022-10-11 | 2022-11-08 | 南通欧惠纺织科技有限公司 | Yarn cutting method for yarn detection |
| CN115311278B (en) * | 2022-10-11 | 2023-12-22 | 南通欧惠纺织科技有限公司 | Yarn segmentation method for yarn detection |
| CN115423807A (en) * | 2022-11-04 | 2022-12-02 | 山东益民服饰有限公司 | Cloth defect detection method based on outlier detection |
| CN115423807B (en) * | 2022-11-04 | 2023-03-24 | 山东益民服饰有限公司 | Cloth defect detection method based on outlier detection |
| CN115618051B (en) * | 2022-12-20 | 2023-03-21 | 楠楠聚智信息科技有限责任公司 | Internet-based smart campus monitoring video storage method |
| CN115618051A (en) * | 2022-12-20 | 2023-01-17 | 楠楠聚智信息科技有限责任公司 | Internet-based smart campus monitoring video storage method |
| CN116805312A (en) * | 2023-08-21 | 2023-09-26 | 青岛时佳汇服装有限公司 | Knitted fabric quality detection method based on image processing |
| CN116805312B (en) * | 2023-08-21 | 2024-01-05 | 青岛时佳汇服装有限公司 | Knitted fabric quality detection method based on image processing |
| CN117152444A (en) * | 2023-10-30 | 2023-12-01 | 山东泰普锂业科技有限公司 | Equipment data acquisition method and system for lithium battery industry |
| CN117152444B (en) * | 2023-10-30 | 2024-01-26 | 山东泰普锂业科技有限公司 | Equipment data acquisition method and system for lithium battery industry |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114926436A (en) | Defect detection method for periodic pattern fabric | |
| CN107833220B (en) | Fabric defect detection method based on deep convolutional neural network and visual saliency | |
| CN113643289B (en) | Fabric surface defect detection method and system based on image processing | |
| CN108364281B (en) | Ribbon edge flaw defect detection method based on convolutional neural network | |
| CN116205919B (en) | Hardware part production quality detection method and system based on artificial intelligence | |
| CN106846339A (en) | A kind of image detecting method and device | |
| CN107038416B (en) | Pedestrian detection method based on binary image improved HOG characteristics | |
| CN107392142B (en) | Method and device for identifying true and false face | |
| CN116777941B (en) | Profile contour detection method and system based on machine vision | |
| CN115330795B (en) | Cloth burr defect detection method | |
| CN104616026B (en) | A kind of monitoring scene type discrimination method towards intelligent video monitoring | |
| CN115661669B (en) | Method and system for monitoring illegal farmland occupancy based on video monitoring | |
| CN115311507B (en) | Building board classification method based on data processing | |
| CN109726649B (en) | Remote sensing image cloud detection method and system and electronic equipment | |
| CN115330767A (en) | Method for identifying production abnormity of corrosion foil | |
| CN112508923B (en) | Weak and small target detection method | |
| CN115797813B (en) | Water environment pollution detection method based on aerial image | |
| CN116664565A (en) | Hidden crack detection method and system for photovoltaic solar cell | |
| CN117541582B (en) | IGBT insulation quality detection method for high-frequency converter | |
| CN103324931A (en) | Method and device for image color binarization | |
| CN115008255B (en) | Tool wear identification method and device for machine tool | |
| JP2009123234A (en) | Object identification method, apparatus and program | |
| CN115861220A (en) | Cold-rolled strip steel surface defect detection method and system based on improved SSD algorithm | |
| CN114554188A (en) | Mobile phone camera detection method and device based on image sensor pixel array | |
| CN117152143B (en) | Reverse osmosis membrane surface quality detection method and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |