CN102645436B - Engineering ceramic grinding surface damage detection method based on grinding grain removal technology - Google Patents
Engineering ceramic grinding surface damage detection method based on grinding grain removal technology Download PDFInfo
- Publication number
- CN102645436B CN102645436B CN201210101537.8A CN201210101537A CN102645436B CN 102645436 B CN102645436 B CN 102645436B CN 201210101537 A CN201210101537 A CN 201210101537A CN 102645436 B CN102645436 B CN 102645436B
- Authority
- CN
- China
- Prior art keywords
- damage
- grinding
- image
- texture
- gray
- 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.)
- Expired - Fee Related
Links
Images
Landscapes
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses an engineering ceramic grinding surface damage detection method based on a grinding texture removal technology, which comprises the following steps of: collecting a sample grinding surface image, analyzing grinding texture features, reconstructing an image after the removal of grinding textures, dividing the image, extracting and selecting damage feature parameters, designing a classifier and finally obtaining a sample grinding surface damage elevation report. The engineering ceramic grinding surface damage detection method has the beneficial effects that as the grinding textures in the digital image are eliminated by adopting a time-frequency analysis method, only surface damage and background exist in the engineering ceramic grinding surface digital image, so that the damage detection is simple and feasible; and the grinding textures in the engineering ceramic grinding surface image of engineering ceramic with the grinding textures are eliminated so as to be convenient for the surface damage detection. By utilizing the detection method, the correct recognition rate of air hole damage in an engineering ceramic grinding surface with the grinding textures is close to 96 percent; the correct recognition rate of crack damage is above 85 percent; the correct recognition rate of scratch damage reaches 80 percent; and under the removal interference of the grinding textures and under the influence of complicated and changeable crushing self shape, the correct recognition rate of crushing damage reaches 70 percent.
Description
Technical field
The invention belongs to engineering ceramics processing field of non destructive testing, be specifically related to a kind of grinding texture digital picture removal technology.
Background technology
At present, engineering ceramic material has been widely used in the various fields such as space flight, nuclear industry, modern medicine, petrochemical complex, but the detection method of its machined surface quality is still similar with metal material, can not effectively reflect the damage status on engineering ceramics surface.School of Mechanical Engineering of University Of Tianjin advanced ceramics and key lab of the process technology Ministry of Education are devoted to the basic research work that engineering ceramics surface damage detects for many years always, taking engineering ceramics theory as basis and in conjunction with experimental study, adopt digital picture recognition technology, set up a set of engineering ceramics surface damage detection system.Detect but adopt digital picture recognition technology to be applied to engineering ceramics surface damage, having technological difficulties is exactly that grinding texture effects on surface damage check in engineering ceramics surface image can cause interference, affects the accuracy rate of damage check.
Summary of the invention
For above-mentioned prior art, the invention provides a kind of engineering ceramics surface damage detection method based on grinding texture removal technology, adopt the method for time frequency analysis to eliminate the grinding texture in digital picture, in the digital picture on gained engineering ceramics surface, only there is surface damage and background, thereby make damage check become simple possible.Realize there being the engineering ceramics of grinding texture, eliminate the grinding texture in engineering ceramics surface image, so that the detection of surface damage.
In order to solve the problems of the technologies described above, the present invention is based on the engineering ceramics surface damage detection method of grinding texture removal technology, comprise the following steps:
Exemplar grinding skin image acquisition step: realize the collection of exemplar grinding skin image, and import computing machine after the image collecting is converted to digitized image;
Grinding texture signature analysis step: utilize gray level co-occurrence matrixes to extract exemplar grinding skin textural characteristics parameter energy and entropy, and judge that with this whether this exemplar grinding skin exists grinding texture, if there is grinding texture, removes grinding texture;
Remove grinding texture step: select Fourier transform, digitized image is transformed into frequency domain from time domain, thereby convert the original digitized image receiving to spectrogram and phase diagram; In spectrogram, adopt frequency domain filter to carry out filtering to spectral image, remove high-energy frequency content corresponding to grinding texture; Finally adopt Fourier inversion to return time domain, obtain removing the reconstructed image after texture;
Image segmentation step, adopt Canny edge detection operator, surface imperfection in above-mentioned reconstructed image is separated from the background of this image, image is cut apart in formation, the former figure computing of dot product obtains cutting apart gray-scale map, and judges whether the result that grinding texture is removed reaches desirable requirement, if F, again spectral image is carried out to filtering by the frequency span that regulates frequency domain filter, until meet the demands;
The extraction of damaging diagnostic parameter and select step, extracts damaging diagnostic parameter from cutting apart image and cutting apart, and therefrom selects damage field circularity, damage field gray average and damage field curvature characteristic parameter gray-scale map;
Classifier design step, utilizes classification tree that exemplar grinding skin is damaged and classified;
Exemplar grinding skin Damage Evaluation step, the ratio of the area of all surface damage field and the total area of surveyed area, calculates damage ratio, and combined with texture characteristic parameter draws the appraisal report of this exemplar grinding skin damage.
Further, in grinding texture signature analysis step, if the threshold values of energy is 0.15～0.20, the threshold values of entropy is 2.3～2.5, show that exemplar grinding skin exists the result of grinding texture.
In image segmentation step, the threshold values of Canny edge detection operator is 0.28～0.35.
In image segmentation step, when grinding texture removal result does not meet ideal conditions, regulating the frequency span scope of frequency domain filter is 4～16 pixels.
In classifier design step, described exemplar grinding skin damage comprises fragmentation, pore, crackle and scratch damage.
In classifier design step, the input using damage field circularity C, damage field gray average MNQ and tri-characteristic parameters of damage field curvature K as sorter; And according to circularity characteristic parameter, the damage of exemplar grinding skin is divided into first class and second largest class, and wherein first class comprises broken and pore damage, second largest class comprises fragmentation, crackle and scratch damage;
In first class, the gray-scale value of broken damage is less than the gray-scale value of pore damage;
In second largest class, the gray-scale value of broken damage is all less than the gray-scale value of scratch damage and Crack Damage; The curvature of scratch damage is less than the curvature of Crack Damage;
The mathematic(al) representation of described damage field circularity C is:
In formula (1), S is the area of damage field, and P is the girth of damage field, and the threshold values of circularity C is 30～40;
The mathematic(al) representation of described damage field gray average MNQ is:
In formula (2), x
ibe each damage grey scale pixel value in damage field, because the gray scale difference value of fragmentation damage and pore damage is very large, so the value of gray average ratio is easier to, wherein, the gray average MNQ that broken damage and pore damage zone are divided is 180～210; Be 150～180 to the gray average MNQ that contains broken damage, Crack Damage or scratch damage;
The mathematic(al) representation of described damage field curvature K is:
In formula (3), ρ is the radius-of-curvature of damage field, is less than 0.001 for the damage field curvature K of Crack Damage and scratch damage.
Compared with prior art, the invention has the beneficial effects as follows:
Not removing before texture, grinding damage is because the interference of grinding texture cannot detect substantially, and the present invention adopts the method for time frequency analysis, and in time domain, the existence of texture feature extraction parameter decision texture is whether; Again imagery exploitation is fourier transformed into frequency domain, the characteristic according to texture in frequency domain, utilizes frequency domain filter filtering grinding texture; Utilize image Fourier inversion, the image after reconstruction filtering, determines image segmentation algorithm and operator according to the characteristic of reconstructed image in time domain.
Apply detection system of the present invention several pictures are carried out to damage check, the result of detection is as shown in table 1.From the result detecting, the recognition correct rate of pore damage reaches approximately 96%; The recognition correct rate of Crack Damage is more than 85%; The recognition correct rate of scratch damage has reached 80%; Even if broken damage too much, and the impact that removed by texture, inevitably cause the loss of damage information, but the discrimination of broken damage is also in 70% left and right.
Brief description of the drawings
Fig. 1 is the engineering ceramics surface damage detection method process flow diagram that the present invention is based on grinding texture removal technology;
Fig. 2 (a) to Fig. 2 (e) be the key step of utilizing the present invention to carry out the detection of engineering ceramics surface damage,
Wherein: Fig. 2 (a) is the former figure of digitizing on engineering ceramics surface,
Fig. 2 (b) is the reconstructed image after texture is removed,
Fig. 2 (c) is cut apart gray-scale map,
Fig. 2 (d) is damage result labeled graph,
Fig. 2 (e) is damage check result schematic diagram;
Fig. 3 is problem and the solution schematic diagram existing after texture is removed;
Fig. 4 is the schematic diagram that utilizes sorter to detect damage in the present invention.
Embodiment
Below in conjunction with embodiment, the present invention is described in further detail.
The main design idea that the present invention is based on the engineering ceramics surface damage detection method of grinding texture removal technology is: the feature that all there will be grinding texture for engineering ceramics surface under general processing conditions, whether texture feature extraction parameter has texture to judge to grinding skin, effectively has or not texture to judge to grinding skin.The impact of the mode filtering grinding texture that utilizes frequency domain filtering on damage check, and according to the situation of practical application, adjust frequency domain filter parameter, in effectively removing texture, reduce the loss of damage information as far as possible.Actual conditions after removing according to texture, are analyzing on the basis of various treatment effects, and integrated use Region Segmentation and edge detecting technology, finally selected to be applicable to the Canny algorithm that image is cut apart, and have obtained satisfied segmentation effect.
Utilization the present invention is based on the engineering ceramics surface damage detection method of grinding texture removal technology, the damage on the engineering ceramics surface that any one has been processed detects, comprise: exemplar grinding skin image acquisition, grinding texture signature analysis, remove that grinding texture, image are cut apart, the extraction of damaging diagnostic parameter and selection, classifier design, finally draw the appraisal report of exemplar grinding skin damage, as shown in Figure 1, concrete steps are as follows:
(1) exemplar grinding skin image acquisition:
Utilize SK-2003 industrial microscope, CCD camera, light source and capture card to realize the collection of exemplar grinding skin image, obtain image comparatively clearly, and by the image converting digital image collecting, then this digitized image is imported to computing machine, as shown in Fig. 2 (a).
(2) grinding texture signature analysis:
Although, utilize the energy that extracts, entropy, contrast, in four textural characteristics parameters of correlativity any one, whether can distinguish easily texture exists, but in order to increase the accuracy of judgement, the present invention gets energy and entropy have or not texture foundation as judgement, utilize gray level co-occurrence matrixes to extract exemplar grinding skin textural characteristics parameter energy and entropy, and judge with this whether this exemplar grinding skin exists grinding texture, if the threshold values of energy is 0.15～0.2, the threshold values of entropy is 2.3～2.5, the threshold values of energy is preferably 0.2, the threshold values of entropy is preferably 2.4, show that exemplar grinding skin exists the result of grinding texture, if there is grinding texture, remove grinding texture.
(3) remove grinding texture:
Select Fourier transform, digitized image is transformed into frequency domain from time domain, thereby convert the original digitized image receiving to spectrogram and phase diagram; In spectrogram, adopt frequency domain filter to carry out filtering to spectral image, remove high-energy frequency content corresponding to grinding texture; Finally adopt Fourier inversion to return time domain, obtain removing the reconstructed image after texture, namely there is no the image of interference of texture, as shown in Fig. 2 (b), it is one of most important link in the present invention that this grinding texture is removed.
There is the image of main texture for a width, high-energy frequency content in its frequency spectrum is all concentrated in the vertical direction of its main texture, if want to remove grinding texture feature, as long as the energy in the vertical direction of the main texture of image is set to 0, and then utilize Fourier inversion to be reconstructed original image.Because Fourier inversion is linear change, therefore the half-tone information of time-domain diagram picture can completely nondestructively remain to frequency domain.Remove textural characteristics if want to strengthen damage defect feature, be beneficial to identification and the classification of damage, the present invention be adopt frequency domain filter to high-energy frequency content in frequency spectrum, suppress filtering.
Also have, for in engineering ceramics surfacing process, produce such as crackle, fragmentation, pore, the surface damages such as cut, from the direction of damage, in most cases, damage defect distributes more even in various angles, and its directivity is not strong, or from another angle, the main grain direction that damage defect direction and Grinding Process produce is inconsistent; From the energy frequency content of damage, the high-energy frequency content of main texture frequency spectrum will be far longer than the energy frequency content of damage defect frequency spectrum, so main grain direction of concentrating by finding out the high-energy frequency content of frequency spectrum, and by its filtering in frequency domain, just can effectively be strengthened damage, and then cut apart and calculate extraction damage characteristic by suitable image.How designing best frequency domain filter is the key that texture is removed, and it directly affects the effect that texture is removed.Determine the concentrated region of high-energy frequency content in spectral image by totalizer, utilized frequency domain filter to carry out filtering processing to the spectrogram of image, utilized the phase diagram of filtered spectral image combining image to carry out Fourier inversion reconstructed image.In reconstructed image, the main texture of grinding is suppressed, and damage defect is effectively strengthened, and then just can adopt threshold segmentation method that damage has been separated from image background.
(4) image is cut apart:
For above-mentioned reconstructed image is separated from the background of this image surface imperfection such as fragmentation, pore, crackle, and then adopt the formation of Canny edge detection operator to cut apart image, the threshold values of Canny edge detection operator is 0.28～0.35, and its preferred threshold values is 0.3; This is cut apart image and carries out the former figure computing of dot product and obtain cutting apart gray-scale map, whether according to this, to cut apart the result that gray-scale map judges that grinding texture removes desirable, if F, and again spectral image is carried out to filtering by the frequency span that regulates frequency domain filter, regulating the frequency span scope of frequency domain filter is 4～16 pixels.Again obtain cutting apart gray-scale map, until meet the demands.As shown in Fig. 2 (c).
As shown in Figure 3, the removal of grinding texture, can produce certain negative influence, as causes the loss of gray scale energy ingredient to make the dimmed loss with causing some damage informations of picture.The effect that the former can cause texture to be removed is difficult to judgement, and the latter can affect testing result.Therefore can utilize image segmentation algorithm and threshold values to select to eliminate above impact.Because the loss of damage information is inevitably, in the present invention, adopts and wave filter is carried out to frequency span parameter adjustment overcome because damage information is lost and caused the mistake of damage classifying below.
(5) extraction of damaging diagnostic parameter and selection:
From above-mentioned cut apart image and cut apart gray-scale map, extract damaging diagnostic parameter, and therefrom select damage field circularity, damage field gray average and damage field curvature characteristic parameter, and label is carried out in the damage defect region of cutting apart in image, as Fig. 2 (d).
(6) classifier design:
According to the damaging diagnostic parameter extracting, utilize classification tree to detect damage, according to the observation and analysis to the different type of impairments of a large amount of pictures, as far as possible few according to characteristic parameter, as far as possible principle accurately of classification, the present invention, according to the damage field circularity of above-mentioned selection, damage field gray average and damage field curvature characteristic parameter, utilizes classification tree that exemplar grinding skin is damaged and classified, and described exemplar grinding skin damage comprises fragmentation, pore, crackle and scratch damage.
As shown in Figure 4, in classifier design step, the input using damage field circularity C, damage field gray average MNQ and tri-characteristic parameters of damage field curvature K as sorter; And according to circularity characteristic parameter, the damage of exemplar grinding skin is divided into first class and second largest class, and wherein first class comprises broken and pore damage, second largest class comprises fragmentation, crackle and scratch damage;
In first class, the gray-scale value of broken damage is less than the gray-scale value of pore damage;
In second largest class, the gray-scale value of broken damage is all less than the gray-scale value of scratch damage and Crack Damage; The curvature of scratch damage is less than the curvature of Crack Damage;
The mathematic(al) representation of described damage field circularity C is:
In formula (1), S is the area of damage field, and P is the girth of damage field, threshold values yxd=30～40 of circularity C;
The mathematic(al) representation of described damage field gray average MNQ is:
In formula (2), x
ibe each damage grey scale pixel value in damage field, because the gray scale difference value of fragmentation damage and pore damage is very large, so the value of gray average ratio is easier to, wherein, the gray average MNQ that broken damage and pore damage zone are divided is hd2=180～210; Be hd1=150～180 to the gray average MNQ that contains broken damage, Crack Damage or scratch damage;
The mathematic(al) representation of described damage field curvature K is:
In formula (3), ρ is the radius-of-curvature of damage field, is less than k=0.001 for the damage field curvature K of Crack Damage and scratch damage.
(7) exemplar grinding skin Damage Evaluation:
Ratio according to the area of all surface damage field with the total area of surveyed area, calculate damage ratio, combined with texture characteristic parameter, exemplar grinding skin is carried out to comprehensive evaluation, thereby obtain the appraisal report of this exemplar grinding skin damage, at the display interface of appraisal report as shown in Fig. 2 (e).
Although in conjunction with figure, invention has been described above; but the present invention is not limited to above-mentioned embodiment; above-mentioned embodiment is only schematic; instead of restrictive; those of ordinary skill in the art is under enlightenment of the present invention; in the situation that not departing from aim of the present invention, can also make a lot of distortion, within these all belong to protection of the present invention.
Claims (5)
1. an engineering ceramics surface damage detection method of removing technology based on grinding texture, is characterized in that, comprises the following steps:
Exemplar grinding skin image acquisition step: realize the collection of exemplar grinding skin image, and import computing machine after the image collecting is converted to digitized image;
Grinding texture signature analysis step: utilize gray level co-occurrence matrixes to extract exemplar grinding skin textural characteristics parameter energy and entropy, and judge that with this whether this exemplar grinding skin exists grinding texture, if there is grinding texture, removes grinding texture;
Remove grinding texture step: select Fourier transform, digitized image is transformed into frequency domain from time domain, thereby convert the original digitized image receiving to spectrogram and phase diagram; In spectrogram, adopt frequency domain filter to carry out filtering to spectral image, remove high-energy frequency content corresponding to grinding texture; Finally adopt Fourier inversion to return time domain, obtain removing the reconstructed image after texture;
Image segmentation step, adopt Canny edge detection operator, surface imperfection in above-mentioned reconstructed image is separated from the background of this image, image is cut apart in formation, the former figure computing of dot product obtains cutting apart gray-scale map, and judges that whether the result that grinding texture is removed reaches desirable requirement, does not reach desirable requirement if cut apart grinding texture removal effect in gray-scale map, again spectral image is carried out to filtering by the frequency span that regulates frequency domain filter, until meet the demands;
The extraction of damaging diagnostic parameter and select step, extracts damaging diagnostic parameter from cutting apart image and cutting apart, and therefrom selects damage field circularity, damage field gray average and damage field curvature characteristic parameter gray-scale map;
Classifier design step, utilizes classification tree that exemplar grinding skin is damaged and classified, and comprising:
Input using damage field circularity C, damage field gray average MNQ and tri-characteristic parameters of damage field curvature K as sorter; And according to circularity characteristic parameter, the damage of exemplar grinding skin is divided into first class and second largest class, and wherein first class comprises broken and pore damage, second largest class comprises fragmentation, crackle and scratch damage;
In first class, the gray-scale value of broken damage is less than the gray-scale value of pore damage;
In second largest class, the gray-scale value of broken damage is all less than the gray-scale value of scratch damage and Crack Damage; The curvature of scratch damage is less than the curvature of Crack Damage;
The mathematic(al) representation of described damage field circularity C is:
In formula (1), S is the area of damage field, and P is the girth of damage field, and the threshold values of circularity C is 30～40;
The mathematic(al) representation of described damage field gray average MNQ is:
In formula (2), x
ibe each damage grey scale pixel value in damage field, wherein, the gray average MNQ that broken damage and pore damage zone are divided is 180～210; Be 150～180 to the gray average MNQ that contains broken damage, Crack Damage or scratch damage;
The mathematic(al) representation of described damage field curvature K is:
In formula (3), ρ is the radius-of-curvature of damage field, is less than 0.001 for the damage field curvature K of Crack Damage and scratch damage;
Exemplar grinding skin Damage Evaluation step, the ratio of the area of all surface damage field and the total area of surveyed area, calculates damage ratio, and combined with texture characteristic parameter draws the appraisal report of this exemplar grinding skin damage.
2. remove according to claim 1 the engineering ceramics surface damage detection method of technology based on grinding texture, it is characterized in that, in grinding texture signature analysis step, if the threshold values of energy is 0.15～0.20, the threshold values of entropy is 2.3～2.5, show that exemplar grinding skin exists the result of grinding texture.
3. the engineering ceramics surface damage detection method of removing according to claim 1 technology based on grinding texture, is characterized in that, in image segmentation step, the threshold values of Canny edge detection operator is 0.28～0.35.
4. remove according to claim 1 the engineering ceramics surface damage detection method of technology based on grinding texture, it is characterized in that, in image segmentation step, when grinding texture removal result does not meet ideal conditions, regulating the frequency span scope of frequency domain filter is 4～16 pixels.
5. the engineering ceramics surface damage detection method of removing according to claim 1 technology based on grinding texture, is characterized in that, in classifier design step, described exemplar grinding skin damage comprises fragmentation, pore, crackle and scratch damage.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210101537.8A CN102645436B (en) | 2012-04-09 | 2012-04-09 | Engineering ceramic grinding surface damage detection method based on grinding grain removal technology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210101537.8A CN102645436B (en) | 2012-04-09 | 2012-04-09 | Engineering ceramic grinding surface damage detection method based on grinding grain removal technology |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102645436A CN102645436A (en) | 2012-08-22 |
| CN102645436B true CN102645436B (en) | 2014-07-09 |
Family
ID=46658392
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210101537.8A Expired - Fee Related CN102645436B (en) | 2012-04-09 | 2012-04-09 | Engineering ceramic grinding surface damage detection method based on grinding grain removal technology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102645436B (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI463109B (en) * | 2012-09-17 | 2014-12-01 | Ind Tech Res Inst | Inspection method for surface texture |
| CN102937592B (en) * | 2012-10-20 | 2014-07-16 | 山东理工大学 | Ceramic radome pore and material loosening defect automatic detection method |
| JP6377459B2 (en) * | 2014-08-29 | 2018-08-22 | 株式会社ディスコ | Wafer inspection method, grinding and polishing equipment |
| CN108090885B (en) * | 2017-12-20 | 2021-08-27 | 百度在线网络技术（北京）有限公司 | Method and apparatus for processing image |
| CN108982544B (en) * | 2018-06-20 | 2020-09-29 | 青岛联合创智科技有限公司 | Method for detecting defective parts of printed circuit board |
| CN112881432B (en) * | 2021-01-12 | 2022-11-29 | 成都泓睿科技有限责任公司 | Method for detecting bottle mouth cracks of liquid glass bottle |
| CN114240920A (en) * | 2021-12-24 | 2022-03-25 | 苏州凌云视界智能设备有限责任公司 | Appearance defect detection method |
| CN114663430A (en) * | 2022-05-18 | 2022-06-24 | 爱科赛智能科技(浙江)有限公司 | PCB surface defect detection method based on frequency domain information double confirmation |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101178367A (en) * | 2007-09-21 | 2008-05-14 | 天津大学 | Ceramic working surface damnification detecting system |
-
2012
- 2012-04-09 CN CN201210101537.8A patent/CN102645436B/en not_active Expired - Fee Related
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101178367A (en) * | 2007-09-21 | 2008-05-14 | 天津大学 | Ceramic working surface damnification detecting system |
Non-Patent Citations (6)
| Title |
|---|
| 先进陶瓷材料表面/亚表面缺陷无损检测;韩雷 等;《组合机床与自动化加工技术》;20071231;全文 * |
| 工程陶瓷磨削加工表面损伤图像检测;程应科 等;《稀有金属材料与工程》;20080131;第37卷;全文 * |
| 朱洪涛 等.陶瓷磨削机理及其对表面/亚表面损伤的影响.《精密制造与自动化》.2004, |
| 程应科 等.工程陶瓷磨削加工表面损伤图像检测.《稀有金属材料与工程》.2008,第37卷 |
| 陶瓷磨削机理及其对表面/亚表面损伤的影响;朱洪涛 等;《精密制造与自动化》;20041231;全文 * |
| 韩雷 等.先进陶瓷材料表面/亚表面缺陷无损检测.《组合机床与自动化加工技术》.2007, |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102645436A (en) | 2012-08-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102645436B (en) | Engineering ceramic grinding surface damage detection method based on grinding grain removal technology | |
| CN105352873B (en) | The characterizing method of shale pore structure | |
| CN101178367B (en) | Ceramic working surface damnification detecting system | |
| Cope et al. | Plant texture classification using gabor co-occurrences | |
| Wang et al. | Research on crack detection algorithm of the concrete bridge based on image processing | |
| CN105447512B (en) | A kind of detection method and device for the beauty defect that essence slightly combines | |
| CN110399875A (en) | A kind of form of general use information extracting method based on deep learning and pixel projection | |
| CN100592323C (en) | Method for identifying fingerprint facing image quality | |
| CN103175844A (en) | Detection method for scratches and defects on surfaces of metal components | |
| CN103729856B (en) | A kind of Fabric Defects Inspection detection method utilizing S-transformation signal extraction | |
| CN103761743A (en) | Solid wood floor surface defect detecting method based on image fusion and division | |
| EP2136318A3 (en) | Image processing apparatus and image processing method | |
| CN103793709A (en) | Cell recognition method and device, and urine analyzer | |
| CN104751474A (en) | Cascade quick image defect segmentation method | |
| Sawant et al. | Currency recognition using image processing and minimum distance classifier technique | |
| CN108830857A (en) | A kind of adaptive Chinese character rubbings image binaryzation partitioning algorithm | |
| Sehad et al. | Gabor filters for degraded document image binarization | |
| Schmugge et al. | Automatic detection of cracks during power plant inspection | |
| CN105512682A (en) | Secret level marking identification method based on Krawtchouk moment and KNN-SMO classifier | |
| Wang et al. | Singular value decomposition combined with wavelet transform for LCD defect detection | |
| Bin et al. | Automatic inspection of surface crack in solar cell images | |
| Mustapha et al. | Crack Detection on Surfaces Using Digital Image Processing | |
| Fakiris et al. | Sonar Class: A MATLAB toolbox for the classification of side scan sonar imagery, using local textural and reverberational characteristics | |
| Bustos‐Pérez et al. | The quantification of surface abrasion on flint stone tools | |
| CN110619273A (en) | Efficient iris recognition method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20140709 Termination date: 20210409 |