CN114486732B - Ceramic tile defect online detection method based on line scanning three-dimension - Google Patents
Ceramic tile defect online detection method based on line scanning three-dimension Download PDFInfo
- Publication number
- CN114486732B CN114486732B CN202111651839.8A CN202111651839A CN114486732B CN 114486732 B CN114486732 B CN 114486732B CN 202111651839 A CN202111651839 A CN 202111651839A CN 114486732 B CN114486732 B CN 114486732B
- Authority
- CN
- China
- Prior art keywords
- tile
- data
- defect
- dimensional
- class
- 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.)
- Active
Links
- 230000007547 defect Effects 0.000 title claims abstract description 171
- 238000001514 detection method Methods 0.000 title claims abstract description 81
- 239000000919 ceramic Substances 0.000 title claims abstract description 41
- 238000000605 extraction Methods 0.000 claims abstract description 35
- 238000006243 chemical reaction Methods 0.000 claims abstract description 13
- 238000000034 method Methods 0.000 claims description 21
- 238000005259 measurement Methods 0.000 claims description 17
- 238000012545 processing Methods 0.000 claims description 11
- 230000002159 abnormal effect Effects 0.000 claims description 10
- 238000006073 displacement reaction Methods 0.000 claims description 8
- 238000001914 filtration Methods 0.000 claims description 8
- 230000000877 morphologic effect Effects 0.000 claims description 6
- 238000004891 communication Methods 0.000 claims 2
- 230000000694 effects Effects 0.000 abstract description 8
- 238000004519 manufacturing process Methods 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 4
- 230000008901 benefit Effects 0.000 description 3
- 238000005070 sampling Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 230000002950 deficient Effects 0.000 description 2
- 230000032798 delamination Effects 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 230000002411 adverse Effects 0.000 description 1
- 238000005034 decoration Methods 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000002310 reflectometry Methods 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/01—Arrangements or apparatus for facilitating the optical investigation
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/95—Investigating the presence of flaws or contamination characterised by the material or shape of the object to be examined
-
- 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
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Signal Processing (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
The invention provides a line scanning three-dimensional tile defect online detection method, which comprises the following steps: acquiring three-dimensional data corresponding to the ceramic tile; the three-dimensional data are obtained by performing line scanning on the ceramic tile, and comprise ceramic tile depth data and ceramic tile gray scale data; performing image Fang Gaocheng-object space elevation conversion and section splicing on the three-dimensional data to obtain spliced characteristic data, and performing tile surface extraction on the spliced characteristic data to obtain tile area data; and carrying out suspected target extraction, suspected target classification and defect target judgment on the tile area data to obtain a tile defect detection result. The on-line detection method for the tile defects based on the line scanning three-dimension can solve the problem of poor detection effect of the tile defects in the prior art, and improves the detection effect of the tile defects.
Description
Technical Field
The invention relates to the technical field of defect detection, in particular to a tile defect online detection method based on line scanning three-dimension.
Background
The ceramic tile is used as an important floor and wall decoration material, and is widely applied to the modern building repair industry, and the quality of the ceramic tile directly influences the appearance, the performance and the service life of the ceramic tile. Along with the increasing market demand, each large ceramic tile manufacturer has stepped into the stage of large-scale flow line production, but more traditional manual detection means are adopted in the quality detection link of ceramic tile products.
The adoption of manual detection not only generates a great deal of labor cost, but also has the problems of low detection speed, difficult complete unification of detection standards and the like; for tile quality inspection personnel, the tile quality inspection personnel need to work in a strong light environment for a long time, eyes are easy to fatigue, and then the judgment of the surface defects of the tile is affected, and in addition, the bad environment of the production line can also cause adverse effects on the physical health of workers.
At present, although some domestic researches on automatic detection of tile defects exist, the technology is basically a two-dimensional image-based detection technology, the detection result is greatly influenced by an image acquisition environment, particularly illumination conditions, and meanwhile, the detection effect on defect types with small influence on the tile reflectivity is poor, so that the technology is basically not applicable to an actual tile production line.
Disclosure of Invention
The invention provides a linear scanning three-dimensional tile defect online detection method, which is used for solving the problem of poor tile defect detection effect in the prior art and improving the tile defect detection effect.
The invention provides a line scanning three-dimensional tile defect online detection method, which comprises the following steps:
acquiring three-dimensional data corresponding to the ceramic tile; the three-dimensional data are obtained by performing line scanning on the ceramic tile, and comprise ceramic tile depth data and ceramic tile gray scale data;
performing image Fang Gaocheng-object space elevation conversion and section splicing on the three-dimensional data to obtain spliced characteristic data, and performing tile surface extraction on the spliced characteristic data to obtain tile area data;
performing suspected target extraction, suspected target classification and defect target judgment on the tile area data to obtain tile defect detection results;
the suspected target extraction includes: performing binarization processing on the tile depth image and the tile gray level image obtained by converting the tile area data, and extracting a suspected target;
the suspected object classification includes: classifying the suspected targets based on the gray features, the size features and the gray contrast features of the suspected targets to obtain classification results;
the defect target determination includes: and extracting a defect target from the classification result based on the difference characteristics between the tile background texture and the tile defect.
According to the line scanning three-dimensional tile defect online detection method provided by the invention, the acquisition of the three-dimensional data corresponding to the tile comprises the following steps:
based on the displacement encoder, controlling the three-dimensional measuring unit to acquire three-dimensional data corresponding to the ceramic tile;
wherein the three-dimensional measuring unit comprises a plurality of sets of three-dimensional measuring sensors, and each set of three-dimensional measuring sensors comprises a three-dimensional camera and a laser.
According to the line scanning three-dimensional tile defect online detection method provided by the invention, the three-dimensional data is subjected to image Fang Gaocheng to object space elevation conversion and section splicing to obtain spliced characteristic data, and tile surface extraction is performed on the spliced characteristic data to obtain tile area data, and the method comprises the following steps:
based on the calibration file, converting the three-dimensional data from an image Fang Gaocheng to an object elevation;
splicing section data in the converted three-dimensional data corresponding to two adjacent groups of three-dimensional measurement units to obtain spliced characteristic data;
and extracting the tile surface from the splicing characteristic data to obtain the tile area data.
According to the line scanning three-dimensional tile defect online detection method provided by the invention, tile surface extraction is carried out on the splicing characteristic data to obtain the tile area data, and the method comprises the following steps:
and on the basis of the tile surface elevation value and the characteristic that gaps exist between adjacent tiles, performing tile surface extraction on the spliced characteristic data to obtain tile area data.
According to the line scanning three-dimensional tile defect online detection method provided by the invention, the suspected target extraction, suspected target classification and defect target judgment are carried out on the tile region data to obtain tile defect detection results, and the method comprises the following steps:
removing isolated abnormal values from the tile region data by adopting a filtering mode, and then converting the tile region data with the removed isolated abnormal values to obtain a tile depth image and a tile gray level image;
performing binarization processing on the tile depth image and the tile gray level image by adopting a global threshold value or a local self-adaptive threshold value method, and extracting a suspected target;
based on the gray level features, the size features and the gray level contrast features of the suspected targets, carrying out self-adaptive classification on the suspected targets by using a clustering algorithm to obtain classification results;
extracting a defect target from the classification result based on the difference characteristics between the tile background texture and the tile defect;
and merging the defect targets corresponding to the tile depth image with the defect targets corresponding to the tile gray level image to obtain the tile defect detection result.
According to the line scanning three-dimensional tile defect online detection method provided by the invention, the self-adaptive classification is carried out on the suspected targets by using a clustering algorithm based on the gray scale features, the size features and the gray scale contrast features of the suspected targets to obtain classification results, and the method comprises the following steps:
for the suspected target, taking the connected areas as a statistics unit, and carrying out statistics on gray scale characteristics, size characteristics and gray scale contrast characteristics on the connected areas one by one;
and based on the similarity of the gray scale features, the size features and the gray scale contrast features of each connected region in the suspected target, carrying out self-adaptive classification on each connected region by using a clustering algorithm to obtain a classification result.
According to the line scanning three-dimensional tile defect online detection method provided by the invention, the defect targets are extracted from the classification result based on the difference characteristics between the tile background textures and the tile defects, and the method comprises the following steps:
based on the classification result, calculating class characteristics corresponding to the classification result class by class; the class feature comprises: at least one of a representative gray scale feature, a representative size feature, a representative gray scale contrast feature, a total area feature of the connected region, a gray scale uniformity feature, a size uniformity feature, a gray scale contrast uniformity feature, and an internal comprehensive similarity feature;
calculating the inter-class difference characteristic value of any two classes of classification results based on the class characteristics corresponding to the classification results;
calculating the difference total characteristic value between the self class and all other classes class by class based on the inter-class difference characteristic value of the arbitrary two class classification results;
determining the class of the potential defect target based on the difference total characteristic value between the self class and all other classes and the total area characteristic of the connected area of the self class;
and extracting the defect targets from all the connected areas in the class where the potential defect targets are located by utilizing the difference characteristics between the tile background textures and the tile defects.
According to the line scanning three-dimensional tile defect online detection method provided by the invention, the inter-class difference characteristic value comprises the following steps: at least one of a representative gray scale feature difference, a representative size feature difference, a representative gray scale contrast feature difference, a total area feature difference of the connected region, a gray scale consistency feature difference, a size consistency feature difference, a gray scale contrast consistency feature difference, and an internal comprehensive similarity feature difference.
The invention provides a line scanning three-dimensional tile defect online detection method, which further comprises the following steps:
classifying the tile defect detection results based on the performance characteristics of the tile defect detection results in the tile depth data and the tile gray data, and counting the occurrence frequency of various tile defects;
the performance characteristics comprise depth deviation numerical characteristics, gray level deviation numerical characteristics and morphological characteristics of the defect target.
According to the tile defect online detection method based on line scanning three-dimension, three-dimension data corresponding to tiles, namely tile depth data and tile gray data, are obtained through line scanning of the tiles, and after image-space-to-object-space conversion, section splicing and tile surface extraction are carried out on the tile data, suspected target extraction, suspected target classification and defect target judgment are carried out, so that tile defect detection results are obtained. By combining the tile depth data and the tile gray data defect detection results, the reliability of the tile defect detection results can be improved. In addition, three-dimensional data corresponding to the ceramic tile is obtained in a line scanning mode, so that the method has the advantage of high data acquisition speed, and the real-time online detection of the ceramic tile defects on the production line can be realized.
Drawings
In order to more clearly illustrate the invention or the technical solutions of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
Fig. 1 is a schematic flow chart of the line scanning three-dimensional tile defect online detection method provided by the invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the present invention more apparent, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is apparent that the described embodiments are some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The line scanning three-dimensional tile defect online detection method based on the invention is described below with reference to fig. 1.
As shown in fig. 1, the present invention provides a line scanning three-dimensional tile defect online detection method, which comprises:
step 110, acquiring three-dimensional data corresponding to the ceramic tile; the three-dimensional data are obtained by performing line scanning on the tile, and the three-dimensional data comprise tile depth data and tile gray scale data.
It can be understood that the three-dimensional data acquisition device is fixedly arranged above the conveyor belt, the tile is positioned on the conveying conveyor belt, the tile is driven to move through the movement of the conveying conveyor belt, and when the tile moves, the three-dimensional data acquisition device performs line scanning on the tile to acquire three-dimensional data of the tile. The three-dimensional data includes depth data and gray data of the tile.
And 120, performing image Fang Gaocheng-object space elevation conversion and section splicing on the three-dimensional data to obtain spliced characteristic data, and performing tile surface extraction on the spliced characteristic data to obtain tile area data.
It will be appreciated that the preliminary processing of three-dimensional data includes, for example, fang Gaocheng to object elevation conversion, cross-section stitching and tile face extraction. The tile surface extraction is to extract the depth data and the gray data corresponding to each tile individually.
Where image Fang Gaocheng is tile depth data acquired by a three-dimensional measurement unit and the object elevation is the actual elevation of the tile.
And 130, performing suspected target extraction, suspected target classification and defect target judgment on the tile area data to obtain a tile defect detection result.
Tile defect detection, including suspected target extraction, suspected target classification and defective target determination. The tile defect detection is firstly carried out in tile depth data and tile gray data respectively, and then detection results in the two data are combined to obtain a final tile defect detection result.
The suspected target extraction includes: removing isolated abnormal values from the tile region data in a filtering mode, converting the tile region data with the isolated abnormal values removed to obtain a tile depth image and a tile gray level image, performing binarization processing on the tile depth image and the tile gray level image, and extracting a suspected target;
the suspected object classification includes: classifying the suspected targets based on the gray features, the size features and the gray contrast features of the suspected targets to obtain classification results;
the defect target determination includes: and extracting a defect target from the classification result based on the difference characteristics between the tile background texture and the tile defect.
In some embodiments, the acquiring three-dimensional data corresponding to the tile includes:
based on the displacement encoder, controlling the three-dimensional measuring unit to acquire three-dimensional data corresponding to the ceramic tile;
wherein the three-dimensional measuring unit comprises a plurality of sets of three-dimensional measuring sensors, and each set of three-dimensional measuring sensors comprises a three-dimensional camera and a laser.
It will be appreciated that line scanning may be achieved based on a three-dimensional data acquisition device comprising a line scanning three-dimensional measurement unit and a high precision displacement encoder.
The line scanning three-dimensional measuring unit consists of a plurality of sets of line scanning three-dimensional measuring sensors and is fixedly arranged on a gate-type bracket crossing the tile conveying belt at a certain horizontal interval; the line scanning three-dimensional measuring sensor consists of a high-speed three-dimensional camera and a line laser.
The three-dimensional camera acquires the elevation of the tile surface corresponding to the laser line by utilizing a laser triangulation principle, the linear scanning three-dimensional measuring unit can acquire three-dimensional depth data and gray data of a single tile cross section through single measurement, and the three-dimensional depth data and the gray data of the whole tile surface can be acquired through continuous equidistant data acquisition under the control of the displacement encoder.
In some embodiments, the performing image Fang Gaocheng to object space elevation conversion and cross section stitching on the three-dimensional data to obtain stitching feature data, and performing tile surface extraction on the stitching feature data to obtain tile region data, including:
based on the calibration file, converting the three-dimensional data from an image Fang Gaocheng to an object elevation;
splicing section data in the converted three-dimensional data corresponding to two adjacent groups of three-dimensional measurement units to obtain spliced characteristic data;
and extracting the tile surface from the splicing characteristic data to obtain the tile area data.
It will be appreciated that the preliminary processing of three-dimensional data includes, for example, fang Gaocheng to object elevation conversion, cross-section stitching and tile face extraction.
The image Fang Gaocheng to object space elevation conversion is realized by calibrating files.
Section stitching, including stitching of transverse section data between adjacent three-dimensional measurement sensors, and stitching between successive section data along the direction of belt movement. The section splicing operation comprises three-dimensional depth data splicing and gray data splicing.
In some embodiments, the tile surface extracting the tile characteristic data to obtain the tile region data includes:
and on the basis of the tile surface elevation value and the characteristic that gaps exist between adjacent tiles, performing tile surface extraction on the spliced characteristic data to obtain tile area data.
It can be understood that tile surface extraction is to extract the depth data and gray data of each tile corresponding region separately from the continuous depth data (i.e. the depth feature data in the splicing feature data) by locating the tile corresponding region according to the feature that the tile surface elevation value is larger than the conveyor plane and gaps exist between adjacent tiles.
In some embodiments, the performing suspected target extraction, suspected target classification, and defect target determination on the tile region data to obtain a tile defect detection result includes:
removing isolated abnormal values from the tile region data by adopting a filtering mode, and then converting the tile region data with the removed isolated abnormal values to obtain a tile depth image and a tile gray level image;
performing binarization processing on the tile depth image and the tile gray level image by adopting a global threshold value or a local self-adaptive threshold value method, and extracting a suspected target;
based on the gray level features, the size features and the gray level contrast features of the suspected targets, carrying out self-adaptive classification on the suspected targets by using a clustering algorithm to obtain classification results;
extracting a defect target from the classification result based on the difference characteristics between the tile background texture and the tile defect;
and merging the defect targets corresponding to the tile depth image with the defect targets corresponding to the tile gray level image to obtain the tile defect detection result.
In some embodiments, the adaptively classifying the suspected target by using a clustering algorithm based on the gray scale feature, the size feature and the gray scale contrast feature of the suspected target to obtain a classification result includes:
for the suspected target, taking the connected areas as a statistics unit, and carrying out statistics on gray scale characteristics, size characteristics and gray scale contrast characteristics on the connected areas one by one;
and based on the similarity of the gray scale features, the size features and the gray scale contrast features of each connected region in the suspected target, carrying out self-adaptive classification on each connected region by using a clustering algorithm to obtain a classification result.
In some embodiments, the extracting defect targets from the classification results based on the difference features between tile background textures and tile defects comprises:
based on the classification result, calculating class characteristics corresponding to the classification result class by class; the class feature comprises: at least one of a representative gray scale feature, a representative size feature, a representative gray scale contrast feature, a total area feature of the connected region, a gray scale uniformity feature, a size uniformity feature, a gray scale contrast uniformity feature, and an internal comprehensive similarity feature;
calculating the difference characteristic value between any two classes based on the class characteristics corresponding to the classification result; wherein the inter-class difference feature value includes: at least one of a representative gray scale feature difference, a representative size feature difference, a representative gray scale contrast feature difference, a total area feature difference of the connected region, a gray scale consistency feature difference, a size consistency feature difference, a gray scale contrast consistency feature difference, and an internal comprehensive similarity feature difference;
calculating the difference total characteristic value between the self class and all other classes class by class based on the inter-class difference characteristic value of the arbitrary two class classification results;
determining the class of the potential defect target based on the difference total characteristic value among the class and all other classes and the total area characteristic of the connected area of the class;
and extracting the defect targets from all the connected areas in the class where the potential defect targets are located by utilizing the difference characteristics between the tile background textures and the tile defects.
It can be understood that the difference of the representative gray scale features is the difference of the representative gray scale features corresponding to the two classification results; the difference of the representative size features, namely the difference of the representative size features corresponding to the two classification results; representing the difference of the gray contrast characteristics, namely the difference of the gray contrast characteristics corresponding to the two classification results; representing the difference of the gray contrast characteristics, namely the difference of the gray contrast characteristics corresponding to the two classification results; other features are similarly different.
Tile defect detection, including suspected target extraction, suspected target classification and defective target determination.
The method comprises the steps of extracting suspected targets, removing isolated abnormal values in tile area data by adopting a filtering mode (such as median filtering or Gaussian filtering), converting the tile area data with the isolated abnormal values removed to obtain a tile depth image and a tile gray level image, and binarizing the tile depth image and the gray level image, wherein the binarization comprises a method based on a global threshold value or a local self-adaptive threshold value, the tile depth image is obtained by converting the tile depth data, and the tile gray level image is obtained by converting the tile gray level data.
The suspected target classification is to adaptively classify the suspected target extracted by binarization processing by using a clustering algorithm according to the similarity of the gray level characteristic, the size characteristic and the gray level contrast characteristic of the target, so as to obtain a classification result.
And judging the defect targets, namely calculating class characteristics and total difference characteristic values between the self class and other classes according to classification results, determining the class of the potential defect targets according to total difference characteristic values between the self class and other classes and total area characteristics of connected areas of the self class, judging all connected areas in the class of the potential defect targets by utilizing the difference characteristics between the background textures of the ceramic tile and the defects of the ceramic tile, merging the defect targets corresponding to the depth images of the ceramic tile with the defect targets corresponding to the gray images of the ceramic tile, and finally determining the detection results of the defects of the ceramic tile.
In some embodiments, the method for online detecting tile defects based on line scanning three-dimensional technology further comprises:
classifying the tile defect detection results based on the performance characteristics of the tile defect detection results in the tile depth data and the tile gray data, and counting the occurrence frequency of various tile defects;
the representation features comprise depth deviation numerical features, gray scale deviation numerical features and morphological features of the defect target.
It can be understood that the defect classification and statistics are performed on detected tile defects, classification is performed according to the performance characteristics of the detected tile defects in the depth data and the gray data, and the occurrence frequency of various defects is counted.
The depth deviation value characteristic of the target defect refers to the depth value of the target defect relative to the size of the tile background, the gray scale deviation value characteristic of the target defect refers to the gray scale value of the target defect relative to the size of the tile background, and the morphological characteristic of the target defect refers to the size and shape characteristic of the target defect.
The performance characteristics include depth deviation value characteristics, gray scale deviation value characteristics, and morphological characteristics of the defect target. Typical defects that can be detected from tile depth data include cracks, edge defects, unfilled corners, pinholes, dirt fall, glaze bubbles, bubble, stripping, shrinkage, delamination, and distortion; the typical defects can be detected from the tile gray data as well, and in addition, tile defects with no obvious change in elevation such as ink dripping, lack of ink, white edges and pitting can be detected.
And combining the appearance characteristics of the defects in the depth data and the gray data, classifying the defects, and recording the times of various types of defects detected in a period.
In other embodiments, the method for online detecting tile defects based on line scanning three-dimensional is as follows:
in the embodiment, three-dimensional data corresponding to the ceramic tile is acquired based on a line scanning three-dimensional data acquisition device, wherein the line scanning three-dimensional data acquisition device comprises a line scanning three-dimensional measurement unit and a high-precision displacement encoder, the three-dimensional measurement unit consists of 3 sets of line scanning three-dimensional measurement sensors, the three-dimensional measurement unit is fixedly arranged on a door-shaped bracket crossing a ceramic tile conveying belt at a certain horizontal interval, and partial overlapping areas exist in measurement ranges of adjacent three-dimensional measurement sensors in the measurement width direction; the high-precision displacement encoder is used for controlling the three-dimensional measuring unit to continuously and equidistantly acquire data.
The three-dimensional measuring sensor comprises a high-speed three-dimensional camera and a line laser, wherein a plurality of sets of three-dimensional measuring sensors share the same line laser, the laser projects laser lines along the width direction of the conveyor belt, the three-dimensional camera utilizes the laser triangulation principle to acquire the elevation of the surface of the ceramic tile corresponding to the laser lines, the three-dimensional depth data and the gray data of a single cross section can be acquired by single measurement of the line scanning three-dimensional measuring unit, and the depth data and the gray data of the surface of the complete ceramic tile can be acquired by continuously and equidistantly acquiring the ceramic tile on the movement conveyor belt under the control of the displacement encoder.
The sampling interval of the line scanning three-dimensional measuring unit in the measuring cross section direction is 0.10mm, the sampling interval in the moving direction of the conveyor belt is 0.10mm, the measuring precision in the elevation direction is 0.05mm, and the on-line detection speed is 60m/min.
And converting the image space elevation value into the object space elevation by the image Fang Gao-range data (namely, depth data) acquired by the line scanning three-dimensional measuring unit through the calibration file. And (3) obtaining the overlapping area (30 mm) of the adjacent sensor in the section measuring direction by calibrating the section elevation data of the object rotated by the image side, and calculating the overlapping area (about 300 pixel points) of the measured value of the adjacent sensor in the section direction by combining the sampling interval of the sensor in the section measuring direction.
Measuring data of overlapping areas of adjacent line scanning three-dimensional measuring sensors, and calculating an average value of the measuring data as data after splicing the overlapping areas; the section measurement values of all adjacent line scanning three-dimensional measurement sensors are gradually spliced from left to right to form complete cross section data; and finally, splicing the complete cross section data along the moving direction of the conveyor belt. The above-described data splicing operation is also required for gradation data.
After the data splicing operation is completed, according to the characteristic that the elevation value of the tile surface is larger than the plane of the conveyor belt and gaps exist between adjacent tiles, the area corresponding to each tile is positioned from continuous depth data, and further the depth data and gray data corresponding to each tile are extracted independently.
And respectively performing defect detection in the tile depth data and the tile gray data, wherein the detection steps comprise suspected target extraction, suspected target classification and defect target judgment.
Firstly, continuously performing median filtering on tile data twice in a 5*5 window, removing discrete isolated outliers in the tile data, and converting tile depth data into a depth image and converting gray data into a gray image after outlier processing is completed.
And respectively binarizing the tile depth image and the gray level image, wherein the binarization method adopted here is a self-adaptive threshold method based on local neighborhood block mean value.
And for the suspected targets extracted by binarization processing, carrying out self-adaptive classification on the suspected targets by using a clustering algorithm according to the similarity of the gray scale features, the size features and the gray scale contrast features of the targets to obtain classification results.
According to the classification result, class characteristics and the difference total characteristic values between the class and other classes are calculated, the class of the potential defect target is determined according to the difference total characteristic values between the class and other classes and the total area characteristics of the connected areas of the class, and the difference characteristics between the background texture of the tile and the tile defects are utilized to judge all the connected areas in the class of the potential defect target, and the defect targets corresponding to the tile depth images are combined with the defect targets corresponding to the tile gray images to finally determine the tile defect detection result.
And classifying the detected tile defects according to the performance characteristics of the detected tile defects in the depth data and the gray data, and counting the occurrence frequency of various defects. The performance characteristics include depth deviation value characteristics, gray scale deviation value characteristics, and morphological characteristics of the defect target.
Typical defects that can be detected from tile depth data include cracks, edge defects, unfilled corners, pinholes, dirt fall, glaze bubbles, bubble, stripping, shrinkage, delamination, and distortion; the typical defects can be detected from the gray data as well, and in addition, tile defects with no significant change in elevation, such as ink dripping, lack of ink, white edges and pitting, can be detected.
And combining the appearance characteristics of the defects in the depth data and the gray data, classifying the defects, and recording the times of various types of defects detected in a period.
In summary, the line scanning three-dimensional tile defect online detection method provided by the invention comprises the following steps: acquiring three-dimensional data corresponding to the ceramic tile; the three-dimensional data are obtained by performing line scanning on the ceramic tile, and comprise ceramic tile depth data and ceramic tile gray scale data; performing image Fang Gaocheng-object space elevation conversion and section splicing on the three-dimensional data to obtain spliced characteristic data, and performing tile surface extraction on the spliced characteristic data to obtain tile area data; and carrying out suspected target extraction, suspected target classification and defect target judgment on the tile area data to obtain a tile defect detection result.
In the linear scanning three-dimensional tile defect online detection method provided by the invention, three-dimensional data corresponding to the tile, namely tile depth data and tile gray data, are obtained by performing linear scanning on the tile, and then, the tile depth data is subjected to image space-to-object space conversion, section splicing and tile surface extraction, and then, suspected target extraction, suspected target classification and defect target judgment, so that a tile defect detection result is obtained. By combining the tile depth data and the tile gray data defect detection results, the reliability of the tile defect detection results can be improved. In addition, three-dimensional data corresponding to the ceramic tile is obtained in a line scanning mode, so that the method has the advantage of high data acquisition speed, and the real-time online detection of the ceramic tile defects on the production line can be realized.
Therefore, the on-line detection method for the tile defects based on the line scanning three-dimension can solve the problem of poor detection effect of the tile defects in the prior art, and improves the detection effect of the tile defects.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention 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 and scope of the technical solutions of the embodiments of the present invention.
Claims (9)
1. The method for detecting the tile defects on line based on line scanning three-dimension is characterized by comprising the following steps of:
acquiring three-dimensional data corresponding to the ceramic tile; the three-dimensional data are obtained by performing line scanning on the ceramic tile, and comprise ceramic tile depth data and ceramic tile gray scale data;
performing image Fang Gaocheng-object space elevation conversion and section splicing on the three-dimensional data to obtain spliced characteristic data, and performing tile surface extraction on the spliced characteristic data to obtain tile area data;
performing suspected target extraction, suspected target classification and defect target judgment on the tile area data to obtain tile defect detection results;
the suspected target extraction includes: performing binarization processing on the tile depth image and the tile gray level image obtained by converting the tile area data, and extracting a suspected target;
the suspected object classification includes: classifying the suspected targets based on the gray features, the size features and the gray contrast features of the suspected targets to obtain classification results;
the defect target determination includes: and extracting a defect target from the classification result based on the difference characteristics between the tile background texture and the tile defect.
2. The method for online detection of tile defects based on line scanning three-dimensions according to claim 1, wherein the acquiring three-dimensional data corresponding to the tile comprises:
based on the displacement encoder, controlling the three-dimensional measuring unit to acquire three-dimensional data corresponding to the ceramic tile;
wherein the three-dimensional measuring unit comprises a plurality of sets of three-dimensional measuring sensors, and each set of three-dimensional measuring sensors comprises a three-dimensional camera and a laser.
3. The method for online detection of tile defects based on line scanning three-dimensions according to claim 2, wherein the step of performing image Fang Gaocheng to object space elevation conversion and cross section stitching on the three-dimensional data to obtain stitching characteristic data, and performing tile surface extraction on the stitching characteristic data to obtain tile area data comprises the steps of:
based on the calibration file, converting the three-dimensional data from an image Fang Gaocheng to an object elevation;
splicing section data in the converted three-dimensional data corresponding to two adjacent groups of three-dimensional measurement units to obtain spliced characteristic data;
and extracting the tile surface from the splicing characteristic data to obtain the tile area data.
4. The method for online detection of tile defects based on line scanning three dimensions according to claim 3, wherein the performing tile surface extraction on the stitching feature data to obtain the tile region data comprises:
and on the basis of the tile surface elevation value and the characteristic that gaps exist between adjacent tiles, performing tile surface extraction on the spliced characteristic data to obtain tile area data.
5. The line scanning three-dimensional tile defect online detection method according to any one of claims 1 to 4, wherein the performing suspected target extraction, suspected target classification and defect target determination on the tile region data to obtain tile defect detection results comprises:
removing isolated abnormal values from the tile region data by adopting a filtering mode, and then converting the tile region data with the removed isolated abnormal values to obtain a tile depth image and a tile gray level image;
performing binarization processing on the tile depth image and the tile gray level image by adopting a global threshold value or a local self-adaptive threshold value method, and extracting a suspected target;
based on the gray level features, the size features and the gray level contrast features of the suspected targets, carrying out self-adaptive classification on the suspected targets by using a clustering algorithm to obtain classification results;
extracting a defect target from the classification result based on the difference characteristics between the tile background texture and the tile defect;
and merging the defect targets corresponding to the tile depth image with the defect targets corresponding to the tile gray level image to obtain the tile defect detection result.
6. The method for online detection of tile defects based on line scanning three dimensions according to claim 5, wherein the classifying the suspected objects adaptively by using a clustering algorithm based on gray features, size features and gray contrast features of the suspected objects to obtain classification results comprises:
taking the communication areas as a statistics unit for the suspected targets, and counting gray scale features, size features and gray scale contrast features of the communication areas;
and based on the similarity of the gray scale features, the size features and the gray scale contrast features of each connected region in the suspected target, carrying out self-adaptive classification on each connected region by using a clustering algorithm to obtain a classification result.
7. The line scan three dimensional based tile defect online detection method of claim 5, wherein the extracting defect targets from the classification results based on the difference features between tile background textures and tile defects comprises:
based on the classification result, calculating class characteristics corresponding to the classification result class by class; the class feature comprises: at least one of a representative gray scale feature, a representative size feature, a representative gray scale contrast feature, a total area feature of the connected region, a gray scale uniformity feature, a size uniformity feature, a gray scale contrast uniformity feature, and an internal comprehensive similarity feature;
calculating the inter-class difference characteristic value of any two classes of classification results based on the class characteristics corresponding to the classification results;
calculating the difference total characteristic value between the self class and all other classes class by class based on the inter-class difference characteristic value of the arbitrary two class classification results;
determining the class of the potential defect target based on the difference total characteristic value between the self class and all other classes and the total area characteristic of the connected area of the self class;
and extracting the defect targets from all the connected areas in the class where the potential defect targets are located by utilizing the difference characteristics between the tile background textures and the tile defects.
8. The line-scan three-dimensional tile defect online detection method according to claim 7, wherein the inter-class difference feature value comprises: at least one of a representative gray scale feature difference, a representative size feature difference, a representative gray scale contrast feature difference, a total area feature difference of the connected region, a gray scale consistency feature difference, a size consistency feature difference, a gray scale contrast consistency feature difference, and an internal comprehensive similarity feature difference.
9. The line-scan three-dimensional tile defect online detection method according to claim 5, further comprising:
classifying the tile defect detection results based on the performance characteristics of the tile defect detection results in the tile depth data and the tile gray data, and counting the occurrence frequency of various tile defects;
the performance characteristics comprise depth deviation numerical characteristics, gray level deviation numerical characteristics and morphological characteristics of the defect target.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111651839.8A CN114486732B (en) | 2021-12-30 | 2021-12-30 | Ceramic tile defect online detection method based on line scanning three-dimension |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111651839.8A CN114486732B (en) | 2021-12-30 | 2021-12-30 | Ceramic tile defect online detection method based on line scanning three-dimension |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114486732A CN114486732A (en) | 2022-05-13 |
| CN114486732B true CN114486732B (en) | 2024-04-09 |
Family
ID=81497362
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111651839.8A Active CN114486732B (en) | 2021-12-30 | 2021-12-30 | Ceramic tile defect online detection method based on line scanning three-dimension |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114486732B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115078384A (en) * | 2022-06-16 | 2022-09-20 | 华侨大学 | Quick detection device of stone material large panel surface pit and crack |
| CN117974666B (en) * | 2024-04-01 | 2024-06-25 | 陕西合阳风动工具有限责任公司 | Quality anomaly detection method for non-circular planetary gear |
Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101871895A (en) * | 2010-05-10 | 2010-10-27 | 重庆大学 | Laser scanning imaging nondestructive inspection method for hot continuous casting blank surface defects |
| CN102288613A (en) * | 2011-05-11 | 2011-12-21 | 北京科技大学 | Surface defect detecting method for fusing grey and depth information |
| CN104458755A (en) * | 2014-11-26 | 2015-03-25 | 吴晓军 | Multi-type material surface defect detection method based on machine vision |
| CN104599283A (en) * | 2015-02-10 | 2015-05-06 | 南京林业大学 | Image depth improvement method for camera height recovery based on depth difference |
| CN104749187A (en) * | 2015-03-25 | 2015-07-01 | 武汉武大卓越科技有限责任公司 | Tunnel lining disease detection device based on infrared temperature field and gray level image |
| CN107633516A (en) * | 2017-09-21 | 2018-01-26 | 武汉武大卓越科技有限责任公司 | A kind of method and apparatus for identifying surface deformation class disease |
| CN107743421A (en) * | 2015-06-11 | 2018-02-27 | 宝洁公司 | Apparatus and method for composition to be applied to surface |
| CN108319920A (en) * | 2018-02-05 | 2018-07-24 | 武汉武大卓越科技有限责任公司 | A kind of pavement strip detection and calculation method of parameters scanning three-dimensional point cloud based on line |
| CN108490000A (en) * | 2018-03-13 | 2018-09-04 | 北京科技大学 | A kind of Bar Wire Product surface defect on-line measuring device and method |
| CN110473187A (en) * | 2019-08-08 | 2019-11-19 | 武汉武大卓越科技有限责任公司 | A kind of line scanning three-dimensional pavement crack extract method of object-oriented |
| CN110569786A (en) * | 2019-09-06 | 2019-12-13 | 中国农业科学院农业资源与农业区划研究所 | fruit tree identification and quantity monitoring method and system based on unmanned aerial vehicle data acquisition |
| CN112348773A (en) * | 2020-09-28 | 2021-02-09 | 歌尔股份有限公司 | Screen defect detection method and device and electronic equipment |
| CN113129265A (en) * | 2021-03-18 | 2021-07-16 | 广东工业大学 | Method and device for detecting surface defects of ceramic tiles and storage medium |
| US11153503B1 (en) * | 2018-04-26 | 2021-10-19 | AI Incorporated | Method and apparatus for overexposing images captured by drones |
| CN113701678A (en) * | 2021-09-18 | 2021-11-26 | 武汉光谷卓越科技股份有限公司 | Road surface flatness detection method based on line scanning three-dimension |
| CN113847884A (en) * | 2021-09-18 | 2021-12-28 | 武汉光谷卓越科技股份有限公司 | Fine three-dimensional measurement and modeling method based on line scanning |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CA2951533C (en) * | 2015-12-10 | 2023-08-08 | Ocean Networks Canada Society | Automated generation of digital elevation models |
-
2021
- 2021-12-30 CN CN202111651839.8A patent/CN114486732B/en active Active
Patent Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101871895A (en) * | 2010-05-10 | 2010-10-27 | 重庆大学 | Laser scanning imaging nondestructive inspection method for hot continuous casting blank surface defects |
| CN102288613A (en) * | 2011-05-11 | 2011-12-21 | 北京科技大学 | Surface defect detecting method for fusing grey and depth information |
| CN104458755A (en) * | 2014-11-26 | 2015-03-25 | 吴晓军 | Multi-type material surface defect detection method based on machine vision |
| CN104599283A (en) * | 2015-02-10 | 2015-05-06 | 南京林业大学 | Image depth improvement method for camera height recovery based on depth difference |
| CN104749187A (en) * | 2015-03-25 | 2015-07-01 | 武汉武大卓越科技有限责任公司 | Tunnel lining disease detection device based on infrared temperature field and gray level image |
| CN107743421A (en) * | 2015-06-11 | 2018-02-27 | 宝洁公司 | Apparatus and method for composition to be applied to surface |
| CN107633516A (en) * | 2017-09-21 | 2018-01-26 | 武汉武大卓越科技有限责任公司 | A kind of method and apparatus for identifying surface deformation class disease |
| CN108319920A (en) * | 2018-02-05 | 2018-07-24 | 武汉武大卓越科技有限责任公司 | A kind of pavement strip detection and calculation method of parameters scanning three-dimensional point cloud based on line |
| CN108490000A (en) * | 2018-03-13 | 2018-09-04 | 北京科技大学 | A kind of Bar Wire Product surface defect on-line measuring device and method |
| US11153503B1 (en) * | 2018-04-26 | 2021-10-19 | AI Incorporated | Method and apparatus for overexposing images captured by drones |
| CN110473187A (en) * | 2019-08-08 | 2019-11-19 | 武汉武大卓越科技有限责任公司 | A kind of line scanning three-dimensional pavement crack extract method of object-oriented |
| CN110569786A (en) * | 2019-09-06 | 2019-12-13 | 中国农业科学院农业资源与农业区划研究所 | fruit tree identification and quantity monitoring method and system based on unmanned aerial vehicle data acquisition |
| CN112348773A (en) * | 2020-09-28 | 2021-02-09 | 歌尔股份有限公司 | Screen defect detection method and device and electronic equipment |
| CN113129265A (en) * | 2021-03-18 | 2021-07-16 | 广东工业大学 | Method and device for detecting surface defects of ceramic tiles and storage medium |
| CN113701678A (en) * | 2021-09-18 | 2021-11-26 | 武汉光谷卓越科技股份有限公司 | Road surface flatness detection method based on line scanning three-dimension |
| CN113847884A (en) * | 2021-09-18 | 2021-12-28 | 武汉光谷卓越科技股份有限公司 | Fine three-dimensional measurement and modeling method based on line scanning |
Non-Patent Citations (5)
| Title |
|---|
| Automated Control of Surface Defects on Ceramic Tiles Using 3D Image Analysis;Andrzej Sioma;Materials;第13卷(第5期);1-13 * |
| Defect detection for end surface of ferrite magnetic tile;Tao, Jiayuan 等;PROCEEDINGS OF SPIE;1-9 * |
| 利用高精度三维测量技术进行路面破损检测;李清泉;邹勤;张德津;;武汉大学学报(信息科学版)(第11期);52-67 * |
| 基于机器视觉元件管脚高度检测***研究;范天海;黄丹平;田建平;于少东;吴志鹏;董娜;;光学技术(第01期);105-112 * |
| 基于神经区域生长瓷砖表面缺陷检测;吴冰;成文俊;马贺敏;;机电工程技术(第12期);79-105 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114486732A (en) | 2022-05-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114486732B (en) | Ceramic tile defect online detection method based on line scanning three-dimension | |
| KR101773791B1 (en) | Method and device for inspecting surfaces of an examined object | |
| JP6620477B2 (en) | Method and program for detecting cracks in concrete | |
| CN109870459B (en) | Track slab crack detection method for ballastless track | |
| CN115375686B (en) | Glass edge flaw detection method based on image processing | |
| Samarawickrama et al. | Matlab based automated surface defect detection system for ceremic tiles using image processing | |
| CN104458764B (en) | Curved uneven surface defect identification method based on large-field-depth stripped image projection | |
| CN103630544A (en) | Online visual detection system | |
| Liu et al. | An automatic system for bearing surface tiny defect detection based on multi-angle illuminations | |
| CN113237889A (en) | Multi-scale ceramic detection method and system | |
| CN111968079B (en) | Three-dimensional pavement crack extraction method based on local extremum of section and segmentation sparsity | |
| CN111539927A (en) | Detection process and algorithm of automobile plastic assembly fastening buckle lack-assembly detection device | |
| EP1467176A1 (en) | Inspection system and method | |
| CN114581805A (en) | Coating roller surface defect detection method adopting 3D line laser profile technology | |
| CN114119483A (en) | Image processing technology-based quality detection method and device for light wallboard for building | |
| CN109622404B (en) | Automatic sorting system and method for micro-workpieces based on machine vision | |
| CN109701890A (en) | Magnetic tile surface defect detection and method for sorting | |
| CN110516725B (en) | Machine vision-based wood board stripe spacing and color detection method | |
| CN111815575A (en) | Bearing steel ball part detection method based on machine vision | |
| CA2962809C (en) | System and method for color scanning a moving article | |
| CN114383522B (en) | Method for measuring surface gap and surface difference of workpiece with reflective difference | |
| Hsu et al. | 3D modeling for steel billet images | |
| CN115014142A (en) | Steel tape scale error measuring method based on machine vision | |
| CN110021027B (en) | Edge cutting point calculation method based on binocular vision | |
| CN111141753A (en) | Ceramic tile surface crack detection method based on machine vision |
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 |