CN116168020B - Leather defect detection method - Google Patents
Leather defect detection method Download PDFInfo
- Publication number
- CN116168020B CN116168020B CN202310431937.3A CN202310431937A CN116168020B CN 116168020 B CN116168020 B CN 116168020B CN 202310431937 A CN202310431937 A CN 202310431937A CN 116168020 B CN116168020 B CN 116168020B
- Authority
- CN
- China
- Prior art keywords
- defect
- edge
- hemming
- parameters
- segmented
- 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.)
- Withdrawn - After Issue
Links
- 230000007547 defect Effects 0.000 title claims abstract description 146
- 239000010985 leather Substances 0.000 title claims abstract description 50
- 238000001514 detection method Methods 0.000 title abstract description 12
- 238000000034 method Methods 0.000 claims abstract description 15
- 238000009957 hemming Methods 0.000 claims description 73
- 230000002950 deficient Effects 0.000 claims description 41
- 239000011159 matrix material Substances 0.000 claims description 34
- 239000013598 vector Substances 0.000 claims description 26
- 230000001419 dependent effect Effects 0.000 claims description 10
- 238000003708 edge detection Methods 0.000 claims description 4
- 239000011324 bead Substances 0.000 claims description 3
- 238000003672 processing method Methods 0.000 abstract 1
- 125000004122 cyclic group Chemical group 0.000 description 3
- 238000007781 pre-processing Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 238000005336 cracking Methods 0.000 description 2
- 230000001788 irregular Effects 0.000 description 2
- 230000032683 aging Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
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
- G06T7/0006—Industrial image inspection using a design-rule based approach
-
- 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)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
The application relates to the technical field of data processing, in particular to a leather defect detection method. The method acquires the gray image data of the leather surface identified by the image identification equipment, further processes and analyzes the acquired data, and is focused on improving the processing method of the acquired data, after determining the defect edge according to the data of the gray image of the leather surface, combining the shape of the defect edge and the gray characteristics of the area near the defect edge to obtain related parameters or coefficients for representing the difference between the opening defect and the hole defect in terms of the shape and the gray characteristics, and obtaining a defect judgment coefficient for distinguishing the two defects by utilizing the obtained related parameters or coefficients, thereby accurately distinguishing the two defect types and improving the judgment accuracy of the opening defect and the hole defect.
Description
Technical Field
The application relates to the technical field of data processing, in particular to a leather defect detection method.
Background
In the leather product processing process, the leather surface defects directly affect the product quality, so that accurate judgment needs to be carried out on the leather surface defects, wherein the leather surface defects comprise opening defects and hole defects, the main cause of the opening defects is that the leather ageing hardness is increased and the leather is opened when the leather is subjected to contact cutting of a sharp instrument, the leather is not damaged under the defect condition, the hole defects are generally the leather defects of a small area caused by improper operation in the leather processing process, and the leather is damaged under the defect condition.
However, when detecting the cracking defect and the hole defect in the leather surface defect types, the edge shapes of the two defects are similar and are closed circular or oblate closed edges, so that the opening defect and the hole defect on the leather cannot be accurately distinguished only by the edge detection method at present, and the leather defect types cannot be accurately judged, so that the corresponding fault types cannot be accurately dealt with.
Disclosure of Invention
The application provides a leather defect detection method, which is used for solving the technical problem that the prior art cannot effectively distinguish leather cracking defects from hole defects, and adopts the following technical scheme:
the application discloses a leather defect detection method, which comprises the following steps of:
acquiring a gray image of the leather surface;
edge detection is carried out on the gray level image of the leather surface, and gradient vectors and gradient vector angles of all defect pixel points on the defect edge are determined;
determining a defective pixel point with a gradient vector angle difference value larger than a set differential angle value between the defective pixel point and an adjacent defective pixel point according to the gradient vector angle of each defective pixel point on the defective edge and taking the defective pixel point as a turning point, and segmenting the defective edge according to the determined turning point to obtain a plurality of segmented edges;
calculating the overall hemming parameters, regional hemming parameters and hemming characteristic coefficients of the segmented edges, and calculating the integral rule degree index of the defect edges;
and calculating a defect judgment coefficient for judging two defects of the opening and hole defects in the leather according to the integral hemming parameters of the segmented edge, the regional hemming parameters and the characteristic coefficients and the integral rule degree index of the defect edge, and finishing defect type judgment by using the defect judgment coefficient.
The beneficial effects of the application are as follows:
according to the method, after the defect edge is determined according to the data of the gray level image on the leather surface, the shape of the defect edge and the gray level characteristic of the area near the defect edge are combined to obtain the relevant parameters or coefficients for representing the difference between the opening defect and the hole defect in terms of the shape and the gray level characteristic, and the obtained relevant parameters or coefficients are utilized to obtain the defect judgment coefficients for distinguishing the two defects, so that the accurate distinction of the two defect types can be completed.
Further, the method for determining the integral hemming parameters of the segmented edge comprises the following steps:
scanning the segment edges by adopting a gray level dependency matrix, and then calculating the integral hemming parameters of the segment edges:
wherein ,for the segment edge global hemming parameters,for the number of dependent pixels obtained when the gray dependent matrix is scanned on the segment edges,is the number of all pixels scanned when the gray-scale dependency matrix is scanned on the segment edge.
Further, the regional hemming parameters of the segmented edge are:
wherein ,for the regional hemming parameters of the segmented edge,for the number of pixels in which the gray-scale dependency matrix has a dependency relationship in a single scan of the segment edge,is the statistical number of all pixel points in the gray level dependency matrix.
Further, the hemming characteristic coefficient of the segmented edge is:
wherein ,for the segmented edge curl characteristic coefficients,representing the number of pixel points at the edge asMeeting region hemming parameters when gray-dependent matrix scanning on segmented edges of a displayAnd (3) defective pixels smaller than the region hemming parameter threshold, wherein n represents the number of defective pixels on the segment edge.
Further, the rule degree index of the defect edge as a whole is:
wherein ,a degree of regularity index indicating the entirety of the edge of the defect,representing the number of segmented edges for the defect edge as a whole.
Further, the defect determination coefficients for determining two defects, namely an opening defect and a hole defect in leather are as follows:
wherein ,as the defect determination coefficient,the whole hemming parameters of the sectional edge;a regional hemming parameter for the segmented edge;characteristic coefficients for the segmented edge bead;is the rule degree index of the whole defect edge.
Further, when the defect judgment coefficient is not smaller than the defect judgment coefficient threshold value, the defect at the moment is considered to be an opening defect, otherwise, the defect at the moment is considered to be a hole defect.
Drawings
FIG. 1 is a flow chart of the leather defect detection method of the present application;
fig. 2 is a schematic diagram of the gray scale dependent moment of the present application.
Detailed Description
The conception of the application is as follows:
after data of a gray level image of the leather surface are obtained and a defect edge is determined, the defect edge is segmented according to the number of obvious turning points of the defect edge to obtain a plurality of segmented edges, then the integral hemming parameters, the regional hemming parameters, the hemming characteristic coefficients and the integral rule degree index of the defect edge of the segmented edge are obtained according to the gray level characteristics of the defect, and the defect judgment coefficients for judging two defects of an opening and a hole in the leather are obtained comprehensively according to the obtained quantities, so that the distinction of the opening defect and the hole defect of the leather according to the shape characteristics of the defect and the gray level characteristics around the defect is completed.
The following describes a leather defect detecting method according to the present application in detail with reference to the accompanying drawings and examples.
Method embodiment:
the embodiment of the leather defect detection method provided by the application has the following specific processes:
step one, acquiring a gray image of the leather surface.
The image acquisition electronic equipment is used for acquiring data of the leather surface gray image obtained by shooting the leather surface image and carrying out graying treatment, and carrying out conventional pretreatment, such as noise reduction treatment, contrast enhancement treatment and the like on the leather surface gray image according to the acquired data.
In this embodiment, the preferred image capturing electronic device is an industrial high-definition camera, and any other feasible image capturing electronic device may be used in other embodiments; and, in this embodiment, after the data of the gray image on the leather surface is obtained, preprocessing such as noise reduction and contrast enhancement is performed on the gray image on the leather surface according to the obtained data, and in other embodiments, preprocessing may not be performed or other preprocessing different from the above may be specifically performed.
And secondly, carrying out edge detection on the gray level image of the leather surface, and determining gradient vectors and gradient vector angles of each defective pixel point on the defective edge.
In the embodiment, the Canny operator is adopted to detect the defects of the gray level image on the leather surface, and the gradient vector of each defective pixel point on the defective edge is correspondingly determined while the defect outline, namely the edge of the defect, is determined.
Since the defect shapes of the opening defect and the hole defect are closed curves, the embodiment records the gradient vector of each defective pixel point on the defect edge in the form of a cyclic sequence to obtain a cyclic gradient sequenceWhere m is the total number of defective pixels.
When the gradient vector of the defect pixel point is a transverse gradient vector, it is specificallyCorresponding gradient vector angleWhen the gradient vector of the defective pixel is a transverse gradient vector, it is specificallyCorresponding gradient vector angleGradient vectors of the oblique defective pixel points areCorresponding gradient vector angle。
And thirdly, segmenting the defective edge according to gradient vector angles of all defective pixel points on the defective edge.
According to cyclic gradient sequencesThe gradient vector angle values of any two adjacent defective pixel points can obtain a differential sequence representing the gradient vector angle difference values of the adjacent defective pixel pointsThe method comprises the following steps:
wherein ,for the first differential angle value,for the i-th differential angle value,the gradient vector angle representing the first defective pixel point,represents the gradient vector angle of the mth and the last defective pixel point,and (3) withThe gradient vector angles of the ith and ith-1 defective pixel points are respectively represented.
The resulting sequences were aligned using statisticsCarrying out statistical analysis to find out a differential angle with a differential angle value larger than the set differential angle value, and recording the differential angle as an extremum to obtain an extremum in the sequenceDefective pixel points corresponding to the extreme valuesThe appearance in the image is the turning point on the defective edge line, and the differential angle value is set as in the embodimentIn other embodiments, the set differential angle value may take other values depending on the defect recognition accuracy requirements.
Dividing the defect edge into a plurality of segment edges according to the determined turning points, and counting the number of the segment edges。
And step four, calculating the whole hemming parameters, the regional hemming parameters and the hemming characteristic coefficients of the segmented edges, and calculating the rule degree index of the whole defective edges.
1. The overall hemming parameters of the segmented edge are calculated.
The embodiment adopts a gray level dependency matrix to complete the calculation of the whole hemming parameters of the segmented edge. The gray-scale dependency matrix is a parameter ofWhereinThe distance from the outermost periphery of the matrix to the center point can represent the order of the matrix, which is,In order to obtain the gray scale neighborhood range between the gray scale of all the pixels in the matrix and the gray scale of the matrix core pixels, if the difference between the gray scale value and the gray scale value of the target pixel at the center is smaller thanIndicating that there is a dependency between these points and the target pixel point, the matrix is shown in fig. 2.
Among the parameters of the gray-scale dependency matrix employed in the present embodiment, parametersAt the end of the line of the,50, analyzing pixels near each segment edge of the defect edge in the image by using the gray-scale dependency matrix, wherein the number of all pixels scanned by the gray-scale dependency matrix near the score segment edge isAnd obtaining a dependency result matrix corresponding to each section of area through analysis.
In addition, the gray-scale dependency matrix is a counting matrix, and repeated scanning exists during counting, so that pixels at the same position are only counted once, pixels with dependency relations near the segment edge are obtained, and the number of highlight pixels with dependency relations in the segment edge of the gray-scale dependency matrix scanning is counted as follows。
Calculating the integral hemming parameters of the segmented edge:
wherein ,for the segment edge global hemming parameters,for the number of dependent pixels obtained when the gray dependent matrix is scanned on the segment edges,is the number of all pixels scanned when the gray-scale dependency matrix is scanned on the segment edge.
The integral hemming parameter reflects the segmented edgeThe overall hemming degree is set, the overall hemming parameter threshold is set to be 0.3, and in other embodiments, the overall hemming parameter threshold can be set to be other values according to the actual situation and the detection accuracy requirement of the defect, and the overall hemming parameter is consideredThe whole hemming degree is high, and the edge of the opening is possibly formed; conversely, the overall hemming parametersThe edge curl is low, and the edge of the hole is possible.
2. The regional hemming parameters of the segmented edges are calculated.
Integral hemming parameters for segmented edgesIs an average quantity and there will be some partial hemming and partial non-hemming of a certain segment of edge in the opening, so a more detailed analysis of the segment edge is required.
Calculating the regional hemming parameters of the segmented edges:
wherein ,for the regional hemming parameters of the segmented edge,for the number of pixels in which the gray-scale dependency matrix has a dependency relationship in a single scan of the segment edge,is the statistical number of all pixel points in the gray level dependency matrix.
Regional hemming parametersThe proportion of pixels having a dependency relationship in the gray scale dependency matrix in the pixels included in the entire matrix during a single scan is shown. Similarly, the area hemming parameter threshold is set to 0.3 in the embodiment, and in other embodiments, the area hemming parameter threshold may be set to other values according to the actual situation and the detection accuracy requirement of the defect, and the area hemming parameter is set toThe edge curling degree in the matrix range is high, and the edge of the opening is more likely to be formed; conversely, regional hemming parametersThe edge curl is low in the time range, and is more likely to be the edge of the hole.
3. The hemming characteristic coefficient of the segmented edge is calculated.
The characteristic coefficients of the sectional edge hemming are as follows:
wherein ,for the segmented edge curl characteristic coefficients,representing the number of pixel points at the edge asMeeting region hemming parameters when gray-dependent matrix scanning on segmented edges of a displayAnd (3) defective pixels smaller than the region hemming parameter threshold, wherein n represents the number of defective pixels on the segment edge.
In this embodiment, the segmented edge hemming characteristic coefficientRepresentation ofThe condition is satisfied when gray-scale dependent matrix scanning is performed on the segment edge with n defective pixel pointsThe proportion of defective pixels in the segment is reflected in the amount of the portion of the segment edge line having a low degree of hemming.
Setting the threshold value of the characteristic coefficient of the curling, setting the threshold value of the whole curling parameter to be 0.2 in the embodiment, taking other values of the threshold value of the whole curling parameter according to the actual situation and the detection accuracy requirement of the defects in other embodiments, and setting the characteristic coefficient of the curlingWhen the edge of the subsection has low curling degree, the part occupies more amount, the part is more obvious in the image, and the edge can be regarded as the edge of the hole; conversely, the hemming characteristic coefficientThe portion of the segmented edge with a low degree of hemming is less occupied and is not sufficiently visible in the image and may be considered as the edge of the opening.
4. And calculating the rule degree index of the whole defect edge.
Among the open and hole defects of leather, there is a more remarkable feature that the open defect generally presents a long strip shape, i.e. generally presents a gap comprising 2 segmented edges, whereas the hole generally presents a higher degree of irregularity in shape, which would constitute a gap from a plurality of segmented edges.
The number of segment edges can be used to obtain the rule degree coefficient at the defect location:
wherein ,rule program for representing defect edge wholeThe degree index of the degree index is set,representing the number of segmented edges for the defect edge as a whole.
Index of degree of regularity for defect edge populationThe degree of the integral rule of the defect edge is reflected, and the smaller the number of times of the larger turning of the edge in one defect area, the more regular the outline of the defect area, and conversely, the more irregular the outline of the defect area.
The value range of the value isWhen (when)When 1, it is considered to be an opening defect,the smaller the edge, the more turns the larger the edge, the more irregular the shape, and the more likely to be hole defects.
And fifthly, judging the defect type according to the determined overall hemming parameters, regional hemming parameters and characteristic coefficients of the segmented edge and the overall rule degree index of the defect edge.
According to the above data, a defect determination coefficient for determining defects of openings and holes in leather is proposed:
wherein ,as the defect determination coefficient,the whole hemming parameters of the sectional edge;a regional hemming parameter for the segmented edge;characteristic coefficients for the segmented edge bead;is the rule degree index of the whole defect edge.
The formula comprehensively analyzes the parameters obtained in the above, sets a defect judgment coefficient threshold, sets the defect judgment coefficient threshold to 8 in the embodiment, and can take other values for the whole hemming parameter threshold according to the actual situation and the requirement on the detection accuracy of the defects in other embodiments.
Defect determination coefficientThe defect at this time is considered to be an opening defect; conversely, the defect determination coefficientThe defect at this time is considered to be a hole defect.
The above embodiments are only for illustrating the technical solution of the present application, and not for limiting the same; although the application has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit of the application and are intended to be included within the scope of the application.
Claims (3)
1. A method for detecting leather defects, comprising the steps of:
acquiring a gray image of the leather surface;
edge detection is carried out on the gray level image of the leather surface, and gradient vectors and gradient vector angles of all defect pixel points on the defect edge are determined;
determining a defective pixel point with a gradient vector angle difference value larger than a set differential angle value between the defective pixel point and an adjacent defective pixel point according to the gradient vector angle of each defective pixel point on the defective edge and taking the defective pixel point as a turning point, and segmenting the defective edge according to the determined turning point to obtain a plurality of segmented edges;
calculating the overall hemming parameters, regional hemming parameters and hemming characteristic coefficients of the segmented edges, and calculating the integral rule degree index of the defect edges;
calculating a defect judgment coefficient for judging two defects of the opening and hole defects in leather according to the integral hemming parameters of the sectional edge, the regional hemming parameters and the hemming characteristic coefficients and the integral rule degree index of the defect edge, and finishing defect type judgment by using the defect judgment coefficient;
the method for determining the integral hemming parameters of the segmented edge comprises the following steps:
scanning the segment edges by adopting a gray level dependency matrix, and then calculating the integral hemming parameters of the segment edges:
wherein ,for the segment edge global hemming parameters +.>For the number of dependent pixels obtained during gray scale dependent matrix scanning on the segment edges, +.>The number of all the pixel points scanned during the gray level dependency matrix scanning on the segment edge;
the regional hemming parameters of the segmented edge are:
wherein ,for the regional hemming parameters of the segmented edges, +.>For the number of pixels with dependency relationship in the single scanning of the gray-scale dependency matrix on the segment edge, < +.>The statistics number of all pixel points in the gray level dependency matrix is used;
the hemming characteristic coefficients of the segmented edges are:
wherein ,for the segmented edge hemming feature coefficient +.>The number of pixels at the edge is +.>Satisfies the region hemming parameter +.>Defective pixel points smaller than the region hemming parameter threshold value, n representing the number of defective pixel points on the segment edge;
the overall rule degree index of the defect edge is:
wherein ,a rule degree index indicating the whole of the defective edge, +.>Representing the number of segmented edges for the defect edge as a whole.
2. The method for detecting leather defects according to claim 1, wherein the defect judging coefficients for judging both of the open and the hole defects in the leather are:
wherein ,for defect determination coefficient, < >>The whole hemming parameters of the sectional edge; />A regional hemming parameter for the segmented edge; />Characteristic coefficients for the segmented edge bead; />Is the rule degree index of the whole defect edge.
3. The method according to claim 1, wherein the defect at the time is considered to be an open defect when the defect determination coefficient is not less than the defect determination coefficient threshold, and otherwise the defect at the time is a hole defect.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310431937.3A CN116168020B (en) | 2023-04-21 | 2023-04-21 | Leather defect detection method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310431937.3A CN116168020B (en) | 2023-04-21 | 2023-04-21 | Leather defect detection method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN116168020A CN116168020A (en) | 2023-05-26 |
CN116168020B true CN116168020B (en) | 2023-08-25 |
Family
ID=86418510
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310431937.3A Withdrawn - After Issue CN116168020B (en) | 2023-04-21 | 2023-04-21 | Leather defect detection method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116168020B (en) |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US4974261A (en) * | 1988-11-15 | 1990-11-27 | Matsushita Electric Works, Ltd. | Optical surface inspection method |
CN1835599A (en) * | 2005-02-07 | 2006-09-20 | 三星电子株式会社 | Method and apparatus for processing a bayer-pattern color digital image signal |
JP2009216539A (en) * | 2008-03-11 | 2009-09-24 | Nippon Steel Corp | Detector for hole/crack defect of belt-shaped object |
EP2787485A1 (en) * | 2013-04-02 | 2014-10-08 | Capex Invest GmbH | Method and device for automatic detection of defects in flexible bodies |
CN107862689A (en) * | 2017-11-21 | 2018-03-30 | 广东工业大学 | Leather surface substantially damaged automatic identifying method and computer-readable recording medium |
CN115631132A (en) * | 2022-09-09 | 2023-01-20 | 国能珠海港务有限公司 | Network training method, defect detection method, device, storage medium and equipment |
-
2023
- 2023-04-21 CN CN202310431937.3A patent/CN116168020B/en not_active Withdrawn - After Issue
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US4974261A (en) * | 1988-11-15 | 1990-11-27 | Matsushita Electric Works, Ltd. | Optical surface inspection method |
CN1835599A (en) * | 2005-02-07 | 2006-09-20 | 三星电子株式会社 | Method and apparatus for processing a bayer-pattern color digital image signal |
JP2009216539A (en) * | 2008-03-11 | 2009-09-24 | Nippon Steel Corp | Detector for hole/crack defect of belt-shaped object |
EP2787485A1 (en) * | 2013-04-02 | 2014-10-08 | Capex Invest GmbH | Method and device for automatic detection of defects in flexible bodies |
CN107862689A (en) * | 2017-11-21 | 2018-03-30 | 广东工业大学 | Leather surface substantially damaged automatic identifying method and computer-readable recording medium |
CN115631132A (en) * | 2022-09-09 | 2023-01-20 | 国能珠海港务有限公司 | Network training method, defect detection method, device, storage medium and equipment |
Non-Patent Citations (1)
Title |
---|
高清整皮表面缺陷自动检测与定位;范大煌;丁磊;邓杰航;;《软件导刊》;第18卷(第12期);第146-149页 * |
Also Published As
Publication number | Publication date |
---|---|
CN116168020A (en) | 2023-05-26 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN113689428B (en) | Mechanical part stress corrosion detection method and system based on image processing | |
CN115294113B (en) | Quality detection method for wood veneer | |
CN114937055B (en) | Image self-adaptive segmentation method and system based on artificial intelligence | |
CN115239735B (en) | Communication cabinet surface defect detection method based on computer vision | |
CN116309537B (en) | Defect detection method for oil stain on surface of tab die | |
CN114549522A (en) | Textile quality detection method based on target detection | |
CN107490582B (en) | Assembly line workpiece detection system | |
CN115294099B (en) | Method and system for detecting hairline defect in steel plate rolling process | |
CN116740058B (en) | Quality detection method for solid state disk matched wafer | |
CN115690108A (en) | Aluminum alloy rod production quality evaluation method based on image processing | |
CN114782329A (en) | Bearing defect damage degree evaluation method and system based on image processing | |
CN113838054A (en) | Mechanical part surface damage detection method based on artificial intelligence | |
CN110689524B (en) | No-reference online image definition evaluation method and system | |
CN116645364B (en) | Alloy steel casting air hole defect detection method based on image data | |
CN116883408B (en) | Integrating instrument shell defect detection method based on artificial intelligence | |
CN115100191A (en) | Metal casting defect identification method based on industrial detection | |
CN115272336A (en) | Metal part defect accurate detection method based on gradient vector | |
CN116993742A (en) | Nickel alloy rolling defect detection method based on machine vision | |
CN117173187A (en) | Intelligent valve fault detection system | |
CN117152129B (en) | Visual detection method and system for surface defects of battery cover plate | |
CN116977230B (en) | Scanning electron microscope image optimization enhancement method | |
CN116168020B (en) | Leather defect detection method | |
CN117011291A (en) | Watch shell quality visual detection method | |
CN117333467A (en) | Image processing-based glass bottle body flaw identification and detection method and system | |
CN114742832B (en) | Welding defect detection method for MiniLED thin plate |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
AV01 | Patent right actively abandoned | ||
AV01 | Patent right actively abandoned | ||
AV01 | Patent right actively abandoned |
Granted publication date: 20230825 Effective date of abandoning: 20230905 |
|
AV01 | Patent right actively abandoned |
Granted publication date: 20230825 Effective date of abandoning: 20230905 |