CN105675626A - Character defect detecting method of tire mold - Google Patents

Character defect detecting method of tire mold Download PDF

Info

Publication number
CN105675626A
CN105675626A CN201610107512.7A CN201610107512A CN105675626A CN 105675626 A CN105675626 A CN 105675626A CN 201610107512 A CN201610107512 A CN 201610107512A CN 105675626 A CN105675626 A CN 105675626A
Authority
CN
China
Prior art keywords
image
character
roi
cad
roi image
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.)
Granted
Application number
CN201610107512.7A
Other languages
Chinese (zh)
Other versions
CN105675626B (en
Inventor
蔡念
陈裕潮
张福
刘根
陈新度
王晗
陈新
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Guangdong University of Technology
Original Assignee
Guangdong University of Technology
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Guangdong University of Technology filed Critical Guangdong University of Technology
Priority to CN201610107512.7A priority Critical patent/CN105675626B/en
Publication of CN105675626A publication Critical patent/CN105675626A/en
Application granted granted Critical
Publication of CN105675626B publication Critical patent/CN105675626B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

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
    • G01N21/84—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
    • G01N21/956—Inspecting patterns on the surface of objects
    • 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
    • G01N21/8851—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
    • G01N2021/8887—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 based on image processing techniques

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)
  • Image Processing (AREA)
  • Image Analysis (AREA)

Abstract

The invention discloses a character defect detecting method of a tire mold. The method comprises the following steps of S1, scanning the tire mold of a tire to be detected, collecting and obtaining a group of images, and obtaining the circular-arc-shaped profile of the outer side of the tire; S2, fitting the circle center and the radius of the circular-arc-shaped profile of the outer side of the tire, and then, positioning an image region of the tire mold of the tire as an ROI (region of interest) image to be tested; S3, classifying the ROI image; S4, selecting different methods for processing according to the classification of the ROI image, and obtaining a CAD (computer aided design) image block matched with each ROI image; S5, performing character recognition, and further performing defect judgment according to the character recognition result; S6, responding to the condition of the character defect existence judging result; returning the operation to execute the fourth step and the fifth step so as to select the judgment result with few character defects as the final result after secondary defect judgment. The character defect detecting method has the advantages that the detection stability is high; the detection cost is low; the detection accuracy is high; the false alarm rate is low; the application range is wide; the character defect detecting method can be widely applied to the field of tire mold detection.

Description

A kind of character defect inspection method of tire-mold
Technical field
The present invention relates to image processing field, particularly relate to a kind of character defect inspection method of tire-mold.
Background technology
Explanation of nouns:
ROI:RegionOfInterest, area-of-interest;
NCC:NormalizedCrossCorrelation normalized crosscorrelation.
In tire-mold is produced, quality testing is particularly important, and wherein, the defects detection of character is quality testingImportant process. At present, in industry, mainly rely on human eye for the defects detection of tire-mold character, but,The noisy environment of plant, a large amount of testings and the high request to quality testing, these all make to comply with merelyDetection mode by human eye is difficult to satisfy the demands. Machine vision completes observation and sentences for human eye with machine exactlyDisconnected, the product quality being usually used in high volume production process detects, or is applied to the hazardous environment that is not suitable for peopleOr the implacable occasion of human eye vision etc., compare artificial vision check, can greatly improve accuracy of detection andSpeed, thus enhance productivity, can also avoid human eye vision to detect deviation and the error brought. MachineVision is widely applied in industrial every field. The character defect of machine vision in industrial productsIn detection, applied on a large scale, character defect comprises the stroke defect of character, the biting and wrongly typed of character.Character defect detecting system, requires to detect undesirable character exactly, will be tested with exactly stroke and lackFall into, bite or the character of wrongly typed. Traditional character defect inspection method based on machine vision is mainly with standard wordSymbol image does template, by extracting some features of character, as shape facility, sets up template, by testing imageMate with standard form, if matching result lower than threshold value, thinks it is defective. This method, easilyIn realization, detect effect better. But, for tire-mold, because output is little, from production efficiencyAngle, can not take Standard Module image as template. Application number is 201510437595.1, and name is calledThe Chinese patent application of the tire fetal membrane surface character defect inspection method based on machine vision, has proposed a kind of logicalCross before template matches testing image and CAD design drawing figure are carried out to polar coordinate transform, with flat type to be detectedImage is as template, and the method that flat type CAD image detects as target, specifically at flat type CADFigure obtains in the image block process matching with ROI image to be measured, is by each ROI image to be measured is enteredRow threshold division, and then after ROI image to be measured being classified by morphology operations, will treat according to classificationSurveying ROI image carries out converting bianry image to after pretreatment. Then the CAD that obtains tire-mold to be detected establishesThe flat type image that meter figure is corresponding, compares ROI to be measured according to this flat type image and ROI picture altitude to be measuredImage carries out convergent-divergent, and then intercepts the image block wide with ROI image block successively, calculates coefficient correlation, phaseThe image block that closes coefficient maximum as the image block matching with ROI image to be measured after, carry out character recognition,Thereby carry out defect judgement according to character identification result. This method testing cost is low and the scope of application is wider, canTo detect rapidly. But this method mainly utilizes the information of the large character of picture the first half to enterRow coupling location, in the time that image the first half does not have character or only has small characters, the result of coupling location can go outMistake, cannot ensure all ROI images to be measured can Obtaining Accurate the CAD image block of coupling, may bringLarger detection error. And in the time of Character segmentation, densely distributed for some, the character that size is smaller,While cutting apart indivedual characters may disconnect or with other Characters Stucks, cause wrong report, bring detection error.
Summary of the invention
In order to solve above-mentioned technical problem, the object of this invention is to provide a kind of character defect inspection of tire-moldSurvey method.
The technical solution adopted for the present invention to solve the technical problems is:
A character defect inspection method for tire-mold, comprises step:
S1, successively to tire film to be detected scan and gather obtain one group of image, and respectively to gatherEvery image process rear acquisition tire outer arc shape profile;
Behind the center of circle and radius of S2, matching tire outer arc shape profile, by polar coordinate transform by wheel to be measuredTire outer arc shape image is converted to flat type testing image, and flat type testing image is carried out to Threshold segmentationAfter, location tire fetal membrane image-region is as ROI image to be measured;
S3, respectively each ROI image is carried out to Threshold segmentation, so by morphology operations by Threshold segmentationAfter ROI image classify, obtain the CAD corresponding to CAD design drawing of tire fetal membrane to be detected simultaneouslyFlat type image;
S4, according to the classification of ROI image, select diverse ways to CAD flat type image and ROI figurePicture is processed, and intercepts and obtain and each ROI image phase after treatment on flat type image after treatmentThe CAD image block of coupling;
S5, each ROI image and the CAD image block that mates with it are carried out to character recognition, and then according toCharacter identification result carries out defect judgement;
S6, there is the situation of character defect in response to judgement, thereby return to execution step S4 and S5 enters againAfter the judgement of row defect, select the less judged result of character defect as final result.
Further, described in described step S3, respectively each ROI image is carried out to Threshold segmentation, and then pass throughThe step that morphology operations is classified the ROI image after Threshold segmentation, specifically comprises:
S31, the initial area of calculating respectively each ROI image and elemental height;
S32, ROI image is carried out to Threshold segmentation, divide foreground area;
S33, obtain default Morphological Structuring Elements, connective mark is corroded and carried out to foreground area, andAfter connected domain being screened according to default screening conditions, according to the connected domain quantity filtering out by ROI imageBe divided into category-A and category-B.
Further, after connected domain being screened according to default screening conditions described in described step S33, according toThe connected domain quantity filtering out is divided into ROI image the step of category-A and category-B, and it is specially:
Connected domain is screened, filter out area and be greater than 1/2 initial area and be highly greater than 1/2 elemental heightConnected domain quantity, if the connected domain quantity filtering out equals 0, ROI image is divided into category-A, if sieveThe connected domain quantity of selecting is greater than 0, ROI image is divided into category-B.
Further, described step S4, comprising:
S41, according to the classification of ROI image, according to the iteration of number of processes, be category-A for ROI imageSituation, select successively processing method one and processing method two to carry out CAD flat type image and ROI imageProcess, the situation that is category-B for ROI image, selects processing method two and processing method three to CAD successivelyFlat type image and ROI image are processed;
S42, according to the aspect ratio of CAD flat type image after treatment and ROI image, ROI image is carried outConvergent-divergent;
S43, on CAD flat type image after treatment successively intercept with convergent-divergent after ROI image with wideImage block, calculates the coefficient correlation of the ROI image after each image block and convergent-divergent, and then by coefficient correlationLarge image block is as the CAD image block matching with this ROI image;
Described processing method one is specially: ROI image carried out carrying out morphology processing after Threshold segmentation, andAccording to the ratio of morphology region area after treatment and initial area, location obtains the character area that comprises characterTerritory, and then ROI image is converted to binary image, CAD flat type image is carried out to Threshold segmentation simultaneouslyAfter be converted to bianry image;
Described processing method two is specially: ROI image and CAD flat type image are carried out to standard deviation filtering,And intercept the latter half of the filtered ROI image of standard deviation;
Described processing method three is specially: adopt canny operator to obtain edge image the conversion of ROI imageFor bianry image, CAD flat type image is carried out being converted to bianry image after Threshold segmentation simultaneously.
Further, described step S1, it is specially:
Successively tire film to be detected is scanned and gathered and obtain one group of image, and respectively to gathered everyOpen image and carry out, after image denoising and Threshold segmentation processing, obtaining tire fetal membrane profile, and then according to contour curvatureDisconnect profile, thereby according to the direction of every section of profile, length and curvature, obtain tire outer arc shape profile.
Further, described step S5, comprising:
S51, CAD image block is carried out carrying out morphology operations after Threshold segmentation, according to default connected domainThreshold value is carried out connected domain screening, and CAD image block is divided into small characters region and large character zone;
S52, according to the classification of ROI image, ROI image is carried out to morphology operations, then according to defaultConnected domain threshold value is carried out connected domain screening, and ROI image is also divided into small characters region and large character zone;
S53, according to the classification of ROI image, the large character zone of ROI image and CAD image block is carried outAfter feature extraction and characteristic matching, carry out defect judgement according to matching result;
S54, for the small characters region of ROI image and CAD image block, carry out, after character recognition, knowingNot Huo get character array be divided into multiple character strings, and then ROI image and CAD image recognition obtainedAfter character string is mated, carry out defect judgement according to matching result.
Further, described step S52, it is specially:
The situation that is category-A for ROI image, carries out after Local threshold segmentation and region growing ROI image,Carry out morphology operations, then carry out connected domain screening according to default connected domain threshold value, by ROI image alsoBe divided into small characters region and large character zone;
The situation that is category-B for ROI image, to ROI image carry out mean filter successively, region growing dividesCut and the negate of binaryzation result after carry out morphology operations, be then communicated with according to default connected domain threshold valueTerritory screening, is also divided into small characters region and large character zone by ROI image.
Further, described step S53, it is specially:
The situation that is category-A for ROI image, using the large character zone of ROI image as template, at CADOn the large character zone of image block, carry out NCC coupling, if matching degree is greater than preset matching threshold value, to ROIThe large character zone of image and CAD image block carries out that morphology subtracts each other successively, difference computing and morphological erosionAfter, judge whether the area in the region obtaining is less than predetermined threshold value, if so, judge that this character exists printing to lackFall into, otherwise, judge that this character printing is correct;
The situation that is category-B for ROI image, adopts canny operator to obtain the large character zone of ROI imageEdge after as template, in the enterprising line search coupling of binary image of the large character zone of CAD image block,If matching degree is less than preset matching threshold value, judge that this character exists printing defects, and misregistration character placeThe centre coordinate in region.
Further, described step S54, comprising:
S541, for the small characters region of ROI image and CAD image block, carry out after character recognition, respectivelyObtain two character arrays, and then each character array is divided into multiple character strings;
S542, each character string to ROI image are mated and are searched in the character array of CAD image blockRope, if mate unsuccessfully, performs step S543, otherwise judges that this character string printing is correct;
After S543, adjustment parameter, re-start character zone and divide, and then the small characters district corresponding to this character stringThe ROI image in territory extracts after character again, and identification obtains a new character string;
S544, in the character array of CAD image block this new character string of match search, if mate unsuccessful,Return to the identification of step S543 repeat character string, matching operation, and judge whether in the searching times of regulationThe match is successful, and if so, judge that this character string printing is correct, and be entered in correct characters array, otherwise,Judge that this character string exists printing defects, and the centre coordinate of misregistration character string region;
S545, each character string to CAD image block are carried out match search in correct characters array, ifMate unsuccessfully, perform step S546, otherwise judge that this character string printing is correct;
S546, on the ROI of this character string corresponding region image, extract after character, identification obtains a verificationCharacter array, and then in array, this character string is carried out to match search checking character, if mate unsuccessful,Judge that this character string exists the defect of biting, otherwise judge that this character string printing is correct.
Further, the flat type image that described in described step S3, CAD design drawing is corresponding is by with belowFormula obtains:
After obtaining CAD design drawing and being carried out Threshold segmentation, be converted to binary map, by CAD design drawingMint-mark image as foreground image, and then the minimum circumscribed circle of this foreground image of matching, and obtaining outside this minimumConnect behind the round center of circle and radius, carry out polar coordinate transform according to the center of circle and the radius that obtain, obtain CAD designThe flat type image of figure.
The invention has the beneficial effects as follows: the character defect inspection method of a kind of tire-mold of the present invention, comprising:S1, tire film to be detected is scanned and gathered obtain one group of image successively, and respectively to gathered everyOpen image and process rear acquisition tire outer arc shape profile; The circle of S2, matching tire outer arc shape profileAfter the heart and radius, by polar coordinate transform, tire outer arc shape image to be measured is converted to flat type and treats mappingPicture, and flat type testing image is carried out after Threshold segmentation, location tire fetal membrane image-region is as to be measuredROI image; S3, respectively each ROI image is carried out to Threshold segmentation, so by morphology operations by thresholdROI image after value is cut apart is classified, and the CAD design drawing that simultaneously obtains tire fetal membrane to be detected is correspondingCAD flat type image; S4, according to the classification of ROI image, select diverse ways to CAD flat typeImage and ROI image are processed, and on flat type image after treatment, intercept acquisition with after treatment everyThe CAD image block that individual ROI image matches; S5, to each ROI image and the CAD that mates with itImage block carries out character recognition, and then carries out defect judgement according to character identification result; S6, deposit in response to judgementIn the situation of character defect, again carry out after defect judgement thereby return to execution step S4 and S5, select wordAccord with the less judged result of defect as final result. This method can detect tire fetal membrane to be detected automaticallyCharacter defect, detects that stability is high, testing cost is low, accuracy in detection is high, rate of false alarm is low and applied widely,Can fast and effeciently detect tire fetal membrane character defect.
Brief description of the drawings
Below in conjunction with drawings and Examples, the invention will be further described.
Fig. 1 is the flow chart of the character defect inspection method of a kind of tire-mold of the present invention.
Detailed description of the invention
With reference to Fig. 1, the invention provides a kind of character defect inspection method of tire-mold, comprise step:
S1, successively to tire film to be detected scan and gather obtain one group of image, and respectively to gatherEvery image process rear acquisition tire outer arc shape profile;
Behind the center of circle and radius of S2, matching tire outer arc shape profile, by polar coordinate transform by wheel to be measuredTire outer arc shape image is converted to flat type testing image, and flat type testing image is carried out to Threshold segmentationAfter, location tire fetal membrane image-region is as ROI image to be measured;
S3, respectively each ROI image is carried out to Threshold segmentation, so by morphology operations by Threshold segmentationAfter ROI image classify, obtain the CAD corresponding to CAD design drawing of tire fetal membrane to be detected simultaneouslyFlat type image;
S4, according to the classification of ROI image, select diverse ways to CAD flat type image and ROI figurePicture is processed, and intercepts and obtain and each ROI image phase after treatment on flat type image after treatmentThe CAD image block of coupling;
S5, each ROI image and the CAD image block that mates with it are carried out to character recognition, and then according toCharacter identification result carries out defect judgement;
S6, there is the situation of character defect in response to judgement, thereby return to execution step S4 and S5 enters againAfter the judgement of row defect, select the less judged result of character defect as final result.
Be further used as preferred embodiment, described in described step S3, respectively each ROI image carried outThreshold segmentation, and then the step of the ROI image after Threshold segmentation being classified by morphology operations, toolBody comprises:
S31, the initial area of calculating respectively each ROI image and elemental height;
S32, ROI image is carried out to Threshold segmentation, divide foreground area;
S33, obtain default Morphological Structuring Elements, connective mark is corroded and carried out to foreground area, andAfter connected domain being screened according to default screening conditions, according to the connected domain quantity filtering out by ROI imageBe divided into category-A and category-B.
Be further used as preferred embodiment, the default screening conditions of basis described in described step S33 are to being communicated withAfter screen in territory, according to the connected domain quantity filtering out, ROI image is divided into the step of category-A and category-B,It is specially:
Connected domain is screened, filter out area and be greater than 1/2 initial area and be highly greater than 1/2 elemental heightConnected domain quantity, if the connected domain quantity filtering out equals 0, ROI image is divided into category-A, if sieveThe connected domain quantity of selecting is greater than 0, ROI image is divided into category-B.
Be further used as preferred embodiment, described step S4, comprising:
S41, according to the classification of ROI image, according to the iteration of number of processes, be category-A for ROI imageSituation, select successively processing method one and processing method two to carry out CAD flat type image and ROI imageProcess, the situation that is category-B for ROI image, selects processing method two and processing method three to CAD successivelyFlat type image and ROI image are processed;
S42, according to the aspect ratio of CAD flat type image after treatment and ROI image, ROI image is carried outConvergent-divergent;
S43, on CAD flat type image after treatment successively intercept with convergent-divergent after ROI image with wideImage block, calculates the coefficient correlation of the ROI image after each image block and convergent-divergent, and then by coefficient correlationLarge image block is as the CAD image block matching with this ROI image;
Described processing method one is specially: ROI image carried out carrying out morphology processing after Threshold segmentation, andAccording to the ratio of morphology region area after treatment and initial area, location obtains the character area that comprises characterTerritory, and then ROI image is converted to binary image, CAD flat type image is carried out to Threshold segmentation simultaneouslyAfter be converted to bianry image;
Described processing method two is specially: ROI image and CAD flat type image are carried out to standard deviation filtering,And intercept the latter half of the filtered ROI image of standard deviation;
Described processing method three is specially: adopt canny operator to obtain edge image the conversion of ROI imageFor bianry image, CAD flat type image is carried out being converted to bianry image after Threshold segmentation simultaneously.
Be further used as preferred embodiment, described step S1, it is specially:
Successively tire film to be detected is scanned and gathered and obtain one group of image, and respectively to gathered everyOpen image and carry out, after image denoising and Threshold segmentation processing, obtaining tire fetal membrane profile, and then according to contour curvatureDisconnect profile, thereby according to the direction of every section of profile, length and curvature, obtain tire outer arc shape profile.
Be further used as preferred embodiment, described step S5, comprising:
S51, CAD image block is carried out carrying out morphology operations after Threshold segmentation, according to default connected domainThreshold value is carried out connected domain screening, and CAD image block is divided into small characters region and large character zone;
S52, according to the classification of ROI image, ROI image is carried out to morphology operations, then according to defaultConnected domain threshold value is carried out connected domain screening, and ROI image is also divided into small characters region and large character zone;
S53, according to the classification of ROI image, the large character zone of ROI image and CAD image block is carried outAfter feature extraction and characteristic matching, carry out defect judgement according to matching result;
S54, for the small characters region of ROI image and CAD image block, carry out, after character recognition, knowingNot Huo get character array be divided into multiple character strings, and then ROI image and CAD image recognition obtainedAfter character string is mated, carry out defect judgement according to matching result.
Be further used as preferred embodiment, described step S52, it is specially:
The situation that is category-A for ROI image, carries out after Local threshold segmentation and region growing ROI image,Carry out morphology operations, then carry out connected domain screening according to default connected domain threshold value, by ROI image alsoBe divided into small characters region and large character zone;
The situation that is category-B for ROI image, to ROI image carry out mean filter successively, region growing dividesCut and the negate of binaryzation result after carry out morphology operations, be then communicated with according to default connected domain threshold valueTerritory screening, is also divided into small characters region and large character zone by ROI image.
Be further used as preferred embodiment, described step S53, it is specially:
The situation that is category-A for ROI image, using the large character zone of ROI image as template, at CADOn the large character zone of image block, carry out NCC coupling, if matching degree is greater than preset matching threshold value, to ROIThe large character zone of image and CAD image block carries out that morphology subtracts each other successively, difference computing and morphological erosionAfter, judge whether the area in the region obtaining is less than predetermined threshold value, if so, judge that this character exists printing to lackFall into, otherwise, judge that this character printing is correct;
The situation that is category-B for ROI image, adopts canny operator to obtain the large character zone of ROI imageEdge after as template, in the enterprising line search coupling of binary image of the large character zone of CAD image block,If matching degree is less than preset matching threshold value, judge that this character exists printing defects, and misregistration character placeThe centre coordinate in region.
Be further used as preferred embodiment, described step S54, comprising:
S541, for the small characters region of ROI image and CAD image block, carry out after character recognition, respectivelyObtain two character arrays, and then each character array is divided into multiple character strings;
S542, each character string to ROI image are mated and are searched in the character array of CAD image blockRope, if mate unsuccessfully, performs step S543, otherwise judges that this character string printing is correct;
After S543, adjustment parameter, re-start character zone and divide, and then the small characters district corresponding to this character stringThe ROI image in territory extracts after character again, and identification obtains a new character string;
S544, in the character array of CAD image block this new character string of match search, if mate unsuccessful,Return to the identification of step S543 repeat character string, matching operation, and judge whether in the searching times of regulationThe match is successful, and if so, judge that this character string printing is correct, and be entered in correct characters array, otherwise,Judge that this character string exists printing defects, and the centre coordinate of misregistration character string region;
S545, each character string to CAD image block are carried out match search in correct characters array, ifMate unsuccessfully, perform step S546, otherwise judge that this character string printing is correct;
S546, on the ROI of this character string corresponding region image, extract after character, identification obtains a verificationCharacter array, and then in array, this character string is carried out to match search checking character, if mate unsuccessful,Judge that this character string exists the defect of biting, otherwise judge that this character string printing is correct.
Be further used as preferred embodiment, described in described step S2, tire outer arc shape profile enteredRow matching is also converted to the step of cutting apart after flat type testing image, and it is specially: matching tire outside circleBehind the center of circle and radius of curved profile, by polar coordinate transform, tire outer arc shape image to be measured is converted toFlat type testing image, and flat type testing image is cut apart.
Be further used as preferred embodiment corresponding straight of CAD design drawing described in described step S3Type image obtains in the following manner:
After obtaining CAD design drawing and being carried out Threshold segmentation, be converted to binary map, by CAD design drawingMint-mark image as foreground image, and then the minimum circumscribed circle of this foreground image of matching, and obtaining outside this minimumConnect behind the round center of circle and radius, carry out polar coordinate transform according to the center of circle and the radius that obtain, obtain CAD designThe flat type image of figure.
Below in conjunction with specific embodiment, the present invention is elaborated.
With reference to Fig. 1, a kind of character defect inspection method of tire-mold, comprises step:
S1, successively to tire film to be detected scan and gather obtain one group of image, and respectively to gatherEvery image process rear acquisition tire outer arc shape profile, it is specially: successively to tire to be detectedFetal membrane scan and gather obtain one group of image, and respectively to every gathered image carry out image denoising andThreshold segmentation obtains tire fetal membrane profile after processing, and then disconnects profile according to contour curvature, thereby according to oftenDirection, length and the curvature of section profile, obtain tire outer arc shape profile.
Disconnect the concrete steps of profile according to contour curvature as follows: judge according to contour area point on profile whetherOn straight line or a camber line, if certain any curvature is consistent with near the curvature of point, represent 2 pointsOn same camber line or same straight line, otherwise, represent that on same camber line or straight line, by 2 points at 2Disconnect. Can obtain direction, length and the curvature of every section of profile by the manner, according to circular arc profileFeature, thus tire outer arc shape profile obtained.
Behind the center of circle and radius of S2, matching tire outer arc shape profile, by polar coordinate transform by wheel to be measuredTire outer arc shape image is converted to flat type testing image, and flat type testing image is carried out to Threshold segmentationAfter, location tire fetal membrane image-region is as ROI image to be measured;
S3, respectively each ROI image is carried out to Threshold segmentation, so by morphology operations by Threshold segmentationAfter ROI image classify, obtain the CAD corresponding to CAD design drawing of tire fetal membrane to be detected simultaneouslyFlat type image;
Wherein, respectively each ROI image is carried out to Threshold segmentation, and then by morphology operations, threshold value is dividedThe step that ROI image after cutting is classified, specifically comprises step S31～S33:
S31, the initial area S_area that calculates respectively each ROI image and elemental height S_Height;
S32, ROI image is carried out to Threshold segmentation, divide foreground area, the threshold value of for example choosing is 120,By the foreground area that is divided into higher than 120;
S33, obtain default Morphological Structuring Elements, connective mark is corroded and carried out to foreground area, andAfter connected domain being screened according to default screening conditions, according to the connected domain quantity filtering out by ROI imageBe divided into category-A and category-B, be specially: obtain default Morphological Structuring Elements, foreground area is corroded and gone forward side by sideThe connective mark of row, and connected domain is screened, filter out area and be greater than 1/2 initial area S_area and heightDegree is greater than the connected domain quantity of 1/2 elemental height S_Height, if the connected domain quantity filtering out equals 0,ROI image is divided into category-A, if the connected domain quantity filtering out is greater than 0, ROI image is divided into category-B.
The flat type image that CAD design drawing is corresponding obtains in the following manner: obtain CAD design drawing alsoCarried out being converted to binary map after Threshold segmentation, using the mint-mark image in CAD design drawing as foreground image,And then the minimum circumscribed circle of this foreground image of matching, and obtain behind the center of circle and radius of this minimum circumscribed circle, according toThe center of circle and the radius that obtain carry out polar coordinate transform, obtain the flat type image of CAD design drawing.
More specifically, the minimum circumscribed circle of this foreground image of matching, and the center of circle and half that obtains this minimum circumscribed circleThe step in footpath, it is specially:
Obtain by non-linear optimum alternative manner matching: the circumscribed circle of this foreground image of matching, according to following formulaBy on edge a little to the summation that adds up of the squared-distance of matching circumscribed circle, and then by outside summation minimumConnect the minimum circumscribed circle of circle as this foreground image, and obtain the center of circle and the radius of this minimum circumscribed circle:
ϵ 2 = Σ i = 1 n ( ( r i - α ) 2 + ( c i - β ) 2 - ρ ) 2
In above formula, ε2Represent the cumulative summation of the squared-distance that arrives a little matching circumscribed circle on edge,(α, β) represents the coordinate in the center of circle, and ρ represents radius of a circle, (ri,ci) represent the coordinate of the point on edge.
Preferably, describedly carry out the step of polar coordinate transform according to the center of circle obtaining and radius, it is specially:
In conjunction with following formula, according to the center of circle and the radius that obtain, CAD design drawing is carried out to polar coordinate transform:
In above formula, (α, β) represents the coordinate of transform center,Represent that the point on CAD design drawing carries outCoordinate after polar coordinate transform, diFor the distance with respect to transform center,For vectorial angle, (ri,ci) be the utmost pointCoordinate before coordinate transform.
S4, according to the classification of ROI image, select diverse ways to CAD flat type image and ROI figurePicture is processed, and intercepts and obtain and each ROI image phase after treatment on flat type image after treatmentThe CAD image block of coupling, specifically comprises step S41～S43:
S41, according to the classification of ROI image, according to the iteration of number of processes, be category-A for ROI imageSituation, select successively processing method one and processing method two to carry out CAD flat type image and ROI imageProcess, the situation that is category-B for ROI image, selects processing method two and processing method three to CAD successivelyFlat type image and ROI image are processed; According to the iteration of number of processes, sequentially select diverse waysCAD flat type image and ROI image are processed, and for example ROI image is category-A, selects for the first timeMethod one is processed, if testing result make mistakes, while processing for the second time, system of selection two;
S42, according to the aspect ratio of CAD flat type image after treatment and ROI image, ROI image is carried outConvergent-divergent;
S43, on CAD flat type image after treatment successively intercept with convergent-divergent after ROI image with wideImage block, calculates the coefficient correlation of the ROI image after each image block and convergent-divergent, and then by coefficient correlationLarge image block is as the CAD image block matching with this ROI image; Coefficient correlation refers to image block and contractingThe coefficient correlation of the matrix of the ROI image after putting.
Described processing method one is specially: ROI image carried out carrying out morphology processing after Threshold segmentation, andAccording to the ratio of morphology region area after treatment and initial area S_area, location obtains and comprises characterCharacter zone, and then ROI image is converted to binary image, CAD flat type image is carried out to threshold simultaneouslyValue is converted to bianry image after cutting apart;
Described processing method two is specially: ROI image and CAD flat type image are carried out to standard deviation filtering,And intercept the latter half of the filtered ROI image of standard deviation;
Described processing method three is specially: adopt canny operator to obtain edge image the conversion of ROI imageFor bianry image, CAD flat type image is carried out being converted to bianry image after Threshold segmentation simultaneously.
S5, each ROI image and the CAD image block that mates with it are carried out to character recognition, and then according toCharacter identification result carries out defect judgement, specifically comprises S51～S54:
S51, CAD image block is carried out carrying out morphology operations after Threshold segmentation, according to default connected domainThreshold value is carried out connected domain screening, and CAD image block is divided into small characters region and large character zone;
S52, according to the classification of ROI image, ROI image is carried out to morphology operations, then according to defaultConnected domain threshold value is carried out connected domain screening, and ROI image is also divided into small characters region and large character zone,Be specially: the situation that is category-A for ROI image, ROI image is carried out to Local threshold segmentation and region lifeAfter length, carry out morphology operations, then carry out connected domain screening according to default connected domain threshold value, by ROIImage is also divided into small characters region and large character zone;
The situation that is category-B for ROI image, to ROI image carry out mean filter successively, region growing dividesCut and the negate of binaryzation result after carry out morphology operations, be then communicated with according to default connected domain threshold valueTerritory screening, is also divided into small characters region and large character zone by ROI image.
S53, according to the classification of ROI image, the large character zone of ROI image and CAD image block is carried outAfter feature extraction and characteristic matching, carry out defect judgement according to matching result, be specially:
The situation that is category-A for ROI image, using the large character zone of ROI image as template, at CADOn the large character zone of image block, carry out NCC coupling, if matching degree is greater than preset matching threshold value, to ROIThe large character zone of image and CAD image block carries out that morphology subtracts each other successively, difference computing and morphological erosionAfter, judge whether the area in the region obtaining is less than predetermined threshold value, if so, judge that this character exists printing to lackFall into, otherwise, judge that this character printing is correct;
The situation that is category-B for ROI image, adopts canny operator to obtain the large character zone of ROI imageSub-pixel precision edge after as template, enterprising at the binary image of the large character zone of CAD image blockLine search coupling, if matching degree is less than preset matching threshold value, judges that this character exists printing defects, and recordThe centre coordinate of error character region.
S54, for the small characters region of ROI image and CAD image block, carry out, after character recognition, knowingNot Huo get character array be divided into multiple character strings, and then ROI image and CAD image recognition obtainedAfter character string is mated, carry out defect judgement according to matching result, specifically comprise step S541～S546:
S541, for the small characters region of ROI image and CAD image block, carry out after character recognition, respectivelyObtain two character arrays, and then each character array is divided into multiple character strings; ROI image and CAD figureBe respectively Array_ROI and Array_CAD as character array corresponding to piece, character string is respectivelyString_ROI[j] and String_CAD[i], wherein, i=1,2,3...N, N is the character that CAD image block is correspondingString number, j=1,2,3...M, M is the character string number that ROI image is corresponding;
S542, each character string String_ROI[j to ROI image], at the character array of CAD image blockIn Array_CAD, carry out match search, if mate unsuccessfully, perform step S543, otherwise judge this wordThe printing of symbol string is correct;
After S543, adjustment parameter, re-start character zone and divide, and then the small characters district corresponding to this character stringThe ROI image in territory extracts after character again, and identification obtains a new character string;
S544, in the character array Array_CAD of CAD image block this new character string of match search,If mate unsuccessfully, return to step S543 repeat character string identification, matching operation, and judge whether on ruleIn fixed searching times, the match is successful, if so, judges that this character string printing is correct, and be entered into correct charactersIn array String_Re, otherwise, judge that this character string exists printing defects, and misregistration character string placeThe centre coordinate in region;
S545, each character string String_CAD[i to CAD image block], in correct characters arrayIn String_Re, carry out match search, if mate unsuccessfully, perform step S546, otherwise judge this characterString printing is correct;
S546, on the ROI of this character string corresponding region image, extract after character, identification obtains a verificationCharacter array String_Recheck, and then checking character in array String_Recheck, this character string to be enteredRow match search, if mate unsuccessfully, judges that this character string exists the defect of biting, otherwise judges this character stringPrinting is correct.
S6, there is the situation of character defect in response to judgement, thereby return to execution step S4 and S5 enters againAfter the judgement of row defect, select the less judged result of character defect as final result.
This method, in testing process, incorporates feedback mechanism, after character defects detection for the first time, if detectedDefective, again carry out character defects detection, after twice defects detection, result less mistake is doneFor final result, improve the stability detecting, reduce wrong rate of false alarm, avoid single testing processThe matching error bringing, has strengthened the degree of accuracy detecting. And for different character classifications, adopt differentDetection method judges, for the character field of the wrong report that easily makes a mistake, incorporates feedback mechanism, for the first timeAfter character zone is divided, the CAD image block and the ROI image that intercept are detected to judgement, if judgement is notQualified, error result is fed back to previous step, change method, carries out character zone for the second time and divides, then heavyNewly detect judgement, improved the degree of accuracy detecting.
Preferably, further comprising the steps of:
S7, successively the ROI image of gathered image is spliced according to acquisition order, obtain measuring wheel to be checkedThe stitching image of the flat type of tire fetal membrane;
S8, the stitching image of flat type is carried out to contrary polar coordinate transform, obtain circular arc stitching image;
S9, to judging defective character, on the correspondence position of circular arc stitching image, highlight, for example,For the existence character of defect of biting, on the correspondence position of circular arc stitching image, use blue circles mark, pinTo there being the character of printing defects, on the correspondence position of circular arc stitching image, use red circle mark.
This method can detect the character defect of tire fetal membrane to be detected automatically, detect stability high, be detected asThis is low and applied widely, can fast and effeciently detect tire fetal membrane character defect.
Be more than that better enforcement of the present invention is illustrated, but the invention is not limited to described realityExecute example, those of ordinary skill in the art also can make all being equal under the prerequisite without prejudice to spirit of the present inventionDistortion or replacement, the modification that these are equal to or replacement are all included in the application's claim limited range.

Claims (10)

1. a character defect inspection method for tire-mold, is characterized in that, comprises step:
S1, tire film to be detected is scanned and gathered obtain one group of image successively, and respectively every gathered image is processed to rear acquisition tire outer arc shape profile;
Behind the center of circle and radius of S2, matching tire outer arc shape profile, by polar coordinate transform, tire outer arc shape image to be measured is converted to flat type testing image, and flat type testing image is carried out after Threshold segmentation, location tire fetal membrane image-region is as ROI image to be measured;
S3, respectively each ROI image is carried out to Threshold segmentation, and then by morphology operations, the ROI image after Threshold segmentation is classified, obtain the CAD flat type image corresponding to CAD design drawing of tire fetal membrane to be detected simultaneously;
S4, according to the classification of ROI image, select diverse ways to process CAD flat type image and ROI image, and on flat type image after treatment, intercept and obtain the CAD image block matching with each ROI image after treatment;
S5, each ROI image and the CAD image block that mates with it are carried out to character recognition, and then carry out defect judgement according to character identification result;
S6, there is the situation of character defect in response to judgement, thereby return to execution step S4 and S5 carries out after defect judgement again, select the less judged result of character defect as final result.
2. the character defect inspection method of a kind of tire-mold according to claim 1, it is characterized in that, described in described step S3, respectively each ROI image is carried out to Threshold segmentation, and then the step of the ROI image after Threshold segmentation being classified by morphology operations, specifically comprise:
S31, the initial area of calculating respectively each ROI image and elemental height;
S32, ROI image is carried out to Threshold segmentation, divide foreground area;
S33, obtain default Morphological Structuring Elements, connective mark is corroded and carried out to foreground area, and after connected domain being screened according to default screening conditions, according to the connected domain quantity filtering out, ROI image is divided into category-A and category-B.
3. the character defect inspection method of a kind of tire-mold according to claim 2, it is characterized in that, after connected domain being screened according to default screening conditions described in described step S33, according to the connected domain quantity filtering out, ROI image is divided into the step of category-A and category-B, it is specially:
Connected domain is screened, filter out the connected domain quantity that area is greater than 1/2 initial area and is highly greater than 1/2 elemental height, if the connected domain quantity filtering out equals 0, ROI image is divided into category-A, if the connected domain quantity filtering out is greater than 0, ROI image is divided into category-B.
4. the character defect inspection method of a kind of tire-mold according to claim 2, is characterized in that, described step S4, comprising:
S41, according to the classification of ROI image, according to the iteration of number of processes, the situation that is category-A for ROI image, select successively processing method one and processing method two to process CAD flat type image and ROI image, the situation that is category-B for ROI image, selects processing method two and processing method three to process CAD flat type image and ROI image successively;
S42, according to the aspect ratio of CAD flat type image after treatment and ROI image, ROI image is carried out to convergent-divergent;
S43, on CAD flat type image after treatment successively intercept with convergent-divergent after ROI image with wide image block, calculate the coefficient correlation of the ROI image after each image block and convergent-divergent, and then using the image block of coefficient correlation maximum as the CAD image block matching with this ROI image;
Described processing method one is specially: ROI image is carried out carrying out morphology processing after Threshold segmentation, and according to the ratio of morphology region area after treatment and initial area, location obtains the character zone that comprises character, and then ROI image is converted to binary image, CAD flat type image is carried out being converted to bianry image after Threshold segmentation simultaneously;
Described processing method two is specially: ROI image and CAD flat type image are carried out to standard deviation filtering, and intercept the latter half of the filtered ROI image of standard deviation;
Described processing method three is specially: adopt canny operator obtain the edge image of ROI image and be converted to bianry image, CAD flat type image is carried out being converted to bianry image after Threshold segmentation simultaneously.
5. the character defect inspection method of a kind of tire-mold according to claim 1, is characterized in that, described step S1, and it is specially:
Successively tire film to be detected is scanned and gathers one group of image of acquisition, and respectively every gathered image is carried out after image denoising and Threshold segmentation processing, obtain tire fetal membrane profile, and then disconnect profile according to contour curvature, thereby according to the direction of every section of profile, length and curvature, obtain tire outer arc shape profile.
6. the character defect inspection method of a kind of tire-mold according to claim 1, is characterized in that, described step S5, comprising:
S51, CAD image block is carried out carrying out morphology operations after Threshold segmentation, carry out connected domain screening according to default connected domain threshold value, CAD image block is divided into small characters region and large character zone;
S52, according to the classification of ROI image, ROI image is carried out to morphology operations, then carry out connected domain screening according to default connected domain threshold value, ROI image is also divided into small characters region and large character zone;
S53, according to the classification of ROI image, the large character zone of ROI image and CAD image block is carried out, after feature extraction and characteristic matching, carrying out defect judgement according to matching result;
S54, for the small characters region of ROI image and CAD image block, carry out after character recognition, the character array that identification is obtained is divided into multiple character strings, and then after the character string of ROI image and the acquisition of CAD image recognition is mated, carries out defect judgement according to matching result.
7. the character defect inspection method of a kind of tire-mold according to claim 6, is characterized in that, described step S52, and it is specially:
The situation that is category-A for ROI image, carries out, after Local threshold segmentation and region growing, carrying out morphology operations to ROI image, then carries out connected domain screening according to default connected domain threshold value, and ROI image is also divided into small characters region and large character zone;
The situation that is category-B for ROI image, to ROI image carry out successively that mean filter, region growing are cut apart and the negate of binaryzation result after carry out morphology operations, then carry out connected domain screening according to default connected domain threshold value, ROI image is also divided into small characters region and large character zone.
8. the character defect inspection method of a kind of tire-mold according to claim 7, is characterized in that, described step S53, and it is specially:
The situation that is category-A for ROI image, using the large character zone of ROI image as template, on the large character zone of CAD image block, carry out NCC coupling, if matching degree is greater than preset matching threshold value, the large character zone of ROI image and CAD image block is carried out successively that morphology subtracts each other, after difference computing and morphological erosion, judges whether the area in the region obtaining is less than predetermined threshold value, if, judge that this character exists printing defects, otherwise, judge that this character printing is correct;
The situation that is category-B for ROI image, adopt canny operator to obtain behind the edge of large character zone of ROI image as template, in the enterprising line search coupling of binary image of the large character zone of CAD image block, if matching degree is less than preset matching threshold value, judge that this character exists printing defects, and the centre coordinate of misregistration character region.
9. the character defect inspection method of a kind of tire-mold according to claim 7, is characterized in that, described step S54, comprising:
S541, for the small characters region of ROI image and CAD image block, carry out after character recognition, obtain respectively two character arrays, and then each character array be divided into multiple character strings;
S542, each character string to ROI image are carried out match search in the character array of CAD image block, if mate unsuccessfully, perform step S543, otherwise judge that this character string printing is correct;
After S543, adjustment parameter, re-start character zone and divide, and then the ROI image in small characters region corresponding to this character string is extracted after character again, identification obtains a new character string;
S544, in the character array of CAD image block this new character string of match search, if mate unsuccessful, return to the identification of step S543 repeat character string, matching operation, and judge whether that the match is successful in the searching times of regulation, if so, judge that this character string printing is correct, and be entered in correct characters array, otherwise, judge that this character string exists printing defects, and the centre coordinate of misregistration character string region;
S545, each character string to CAD image block are carried out match search in correct characters array, if mate unsuccessfully, perform step S546, otherwise judge that this character string printing is correct;
S546, on the ROI of this character string corresponding region image, extract after character, identification obtains the array of checking character, and then in array, this character string is carried out to match search checking character, if mate unsuccessful, judge that this character string exists the defect of biting, otherwise judge that this character string printing is correct.
10. the character defect inspection method of a kind of tire-mold according to claim 1, is characterized in that, the flat type image that described in described step S3, CAD design drawing is corresponding obtains in the following manner:
After obtaining CAD design drawing and being carried out Threshold segmentation, be converted to binary map, using the mint-mark image in CAD design drawing as foreground image, and then the minimum circumscribed circle of this foreground image of matching, and obtain behind the center of circle and radius of this minimum circumscribed circle, carry out polar coordinate transform according to the center of circle and the radius that obtain, obtain the flat type image of CAD design drawing.
CN201610107512.7A 2016-02-26 2016-02-26 A kind of character defect inspection method of tire-mold Active CN105675626B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610107512.7A CN105675626B (en) 2016-02-26 2016-02-26 A kind of character defect inspection method of tire-mold

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610107512.7A CN105675626B (en) 2016-02-26 2016-02-26 A kind of character defect inspection method of tire-mold

Publications (2)

Publication Number Publication Date
CN105675626A true CN105675626A (en) 2016-06-15
CN105675626B CN105675626B (en) 2018-08-07

Family

ID=56305136

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610107512.7A Active CN105675626B (en) 2016-02-26 2016-02-26 A kind of character defect inspection method of tire-mold

Country Status (1)

Country Link
CN (1) CN105675626B (en)

Cited By (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106355572A (en) * 2016-08-19 2017-01-25 广东工业大学 Method for automatic positioning of tire mold image
CN106897990A (en) * 2016-08-31 2017-06-27 广东工业大学 The character defect inspection method of tire-mold
CN107328793A (en) * 2017-06-30 2017-11-07 航天新长征大道科技有限公司 A kind of ornaments surface word print flaw detection method and device based on machine vision
CN107948464A (en) * 2017-09-15 2018-04-20 兰州交通大学 A kind of geometric correction method and system of the laterally offset of printed matter detection image
CN108711148A (en) * 2018-05-11 2018-10-26 沈阳理工大学 A kind of wheel tyre defect intelligent detecting method based on deep learning
CN109142383A (en) * 2018-08-10 2019-01-04 惠州太初科技有限公司 One kind being based on morphologic character defect inspection method and device
CN109685070A (en) * 2019-01-11 2019-04-26 上海大学(浙江·嘉兴)新兴产业研究院 A kind of image pre-processing method
CN109712162A (en) * 2019-01-18 2019-05-03 珠海博明视觉科技有限公司 A kind of cable character defect inspection method and device based on projection histogram difference
CN110088770A (en) * 2016-12-28 2019-08-02 欧姆龙健康医疗事业株式会社 Terminal installation
CN110119459A (en) * 2018-01-24 2019-08-13 纬创资通股份有限公司 Image data retrieval method and image data retrieving apparatus
CN111060527A (en) * 2019-12-30 2020-04-24 歌尔股份有限公司 Character defect detection method and device
CN112966746A (en) * 2020-11-20 2021-06-15 扬州大学 Stable variable gray template generation method suitable for tire defect detection
CN114120310A (en) * 2022-01-27 2022-03-01 廊坊易砚领创科技有限公司 Detection method of tire mold side plate
CN115497099A (en) * 2022-09-23 2022-12-20 神州数码系统集成服务有限公司 Single character image matching and identifying method based on circular scanning
CN116758045A (en) * 2023-07-05 2023-09-15 日照鲁光电子科技有限公司 Surface defect detection method and system for semiconductor light-emitting diode

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2014238292A (en) * 2013-06-06 2014-12-18 株式会社ブリヂストン Appearance inspection device and appearance inspection method
US9110032B2 (en) * 2013-03-14 2015-08-18 Integro Technologies Corp. System and methods for inspecting tire wheel assemblies
CN105067638A (en) * 2015-07-22 2015-11-18 广东工业大学 Tire fetal-membrane surface character defect detection method based on machine vision
CN204842266U (en) * 2015-07-30 2015-12-09 昆山德玛驰自动化设备科技有限公司 CCD character detects machine
CN105283750A (en) * 2013-06-13 2016-01-27 米其林企业总公司 Method for processing a digital image of the surface of a tire in order to detect an anomaly
CN105352974A (en) * 2015-11-13 2016-02-24 广东工业大学 A device for recognizing and detecting surface words of a tyre sidewall of a tyre mold

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9110032B2 (en) * 2013-03-14 2015-08-18 Integro Technologies Corp. System and methods for inspecting tire wheel assemblies
JP2014238292A (en) * 2013-06-06 2014-12-18 株式会社ブリヂストン Appearance inspection device and appearance inspection method
CN105283750A (en) * 2013-06-13 2016-01-27 米其林企业总公司 Method for processing a digital image of the surface of a tire in order to detect an anomaly
CN105067638A (en) * 2015-07-22 2015-11-18 广东工业大学 Tire fetal-membrane surface character defect detection method based on machine vision
CN204842266U (en) * 2015-07-30 2015-12-09 昆山德玛驰自动化设备科技有限公司 CCD character detects machine
CN105352974A (en) * 2015-11-13 2016-02-24 广东工业大学 A device for recognizing and detecting surface words of a tyre sidewall of a tyre mold

Cited By (24)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106355572A (en) * 2016-08-19 2017-01-25 广东工业大学 Method for automatic positioning of tire mold image
CN106355572B (en) * 2016-08-19 2019-05-21 广东工业大学 The automatic positioning method of tire-mold image
CN106897990A (en) * 2016-08-31 2017-06-27 广东工业大学 The character defect inspection method of tire-mold
CN110088770B (en) * 2016-12-28 2023-07-07 欧姆龙健康医疗事业株式会社 Terminal device
CN110088770A (en) * 2016-12-28 2019-08-02 欧姆龙健康医疗事业株式会社 Terminal installation
CN107328793A (en) * 2017-06-30 2017-11-07 航天新长征大道科技有限公司 A kind of ornaments surface word print flaw detection method and device based on machine vision
CN107948464B (en) * 2017-09-15 2019-07-23 兰州交通大学 A kind of geometric correction method and system of the laterally offset of printed matter detection image
CN107948464A (en) * 2017-09-15 2018-04-20 兰州交通大学 A kind of geometric correction method and system of the laterally offset of printed matter detection image
CN110119459A (en) * 2018-01-24 2019-08-13 纬创资通股份有限公司 Image data retrieval method and image data retrieving apparatus
CN108711148B (en) * 2018-05-11 2022-05-27 沈阳理工大学 Tire defect intelligent detection method based on deep learning
CN108711148A (en) * 2018-05-11 2018-10-26 沈阳理工大学 A kind of wheel tyre defect intelligent detecting method based on deep learning
CN109142383A (en) * 2018-08-10 2019-01-04 惠州太初科技有限公司 One kind being based on morphologic character defect inspection method and device
CN109685070A (en) * 2019-01-11 2019-04-26 上海大学(浙江·嘉兴)新兴产业研究院 A kind of image pre-processing method
CN109712162B (en) * 2019-01-18 2023-03-21 珠海博明视觉科技有限公司 Cable character defect detection method and device based on projection histogram difference
CN109712162A (en) * 2019-01-18 2019-05-03 珠海博明视觉科技有限公司 A kind of cable character defect inspection method and device based on projection histogram difference
CN111060527B (en) * 2019-12-30 2021-10-29 歌尔股份有限公司 Character defect detection method and device
CN111060527A (en) * 2019-12-30 2020-04-24 歌尔股份有限公司 Character defect detection method and device
US12002198B2 (en) 2019-12-30 2024-06-04 Goertek Inc. Character defect detection method and device
CN112966746A (en) * 2020-11-20 2021-06-15 扬州大学 Stable variable gray template generation method suitable for tire defect detection
CN114120310A (en) * 2022-01-27 2022-03-01 廊坊易砚领创科技有限公司 Detection method of tire mold side plate
CN115497099A (en) * 2022-09-23 2022-12-20 神州数码系统集成服务有限公司 Single character image matching and identifying method based on circular scanning
CN115497099B (en) * 2022-09-23 2023-04-18 神州数码系统集成服务有限公司 Single character image matching and identifying method based on circular scanning
CN116758045A (en) * 2023-07-05 2023-09-15 日照鲁光电子科技有限公司 Surface defect detection method and system for semiconductor light-emitting diode
CN116758045B (en) * 2023-07-05 2024-01-23 日照鲁光电子科技有限公司 Surface defect detection method and system for semiconductor light-emitting diode

Also Published As

Publication number Publication date
CN105675626B (en) 2018-08-07

Similar Documents

Publication Publication Date Title
CN105675626A (en) Character defect detecting method of tire mold
CN109377485B (en) Machine vision detection method for instant noodle packaging defects
CN107392232B (en) Flotation working condition classification method and system
CN105067638B (en) Tire fetal membrane face character defect inspection method based on machine vision
CN112818988B (en) Automatic identification reading method and system for pointer instrument
CN112699876B (en) Automatic reading method for various meters of gas collecting station
CN109859164B (en) Method for visual inspection of PCBA (printed circuit board assembly) through rapid convolutional neural network
CN105389586A (en) Method for automatically detecting integrity of shrimp body based on computer vision
CN107610114A (en) Optical satellite remote sensing image cloud snow mist detection method based on SVMs
CN111539330B (en) Transformer substation digital display instrument identification method based on double-SVM multi-classifier
CN104483243B (en) Adhered-rice detection and segmentation method, device and system
CN104198497A (en) Surface defect detection method based on visual saliency map and support vector machine
CN111914902B (en) Traditional Chinese medicine identification and surface defect detection method based on deep neural network
CN109003275A (en) The dividing method of weld defect image
CN104537367A (en) VIN code checking method
CN108764312A (en) DS-based optimized multi-index dam defect image detection method
CN113435460A (en) Method for identifying brilliant particle limestone image
CN114926410A (en) Method for detecting appearance defects of brake disc
CN108932471B (en) Vehicle detection method
CN114549493A (en) Magnetic core defect detection system and method based on deep learning
CN114662594B (en) Target feature recognition analysis system
TWI543117B (en) Method for recognizing and locating object
CN109543498A (en) A kind of method for detecting lane lines based on multitask network
CN105956592B (en) A kind of Aircraft Targets detection method based on saliency and SVM
CN117237747B (en) Hardware defect classification and identification method based on artificial intelligence

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant