CN103868924A - Bearing appearance defect detecting algorithm based on visual sense - Google Patents
Bearing appearance defect detecting algorithm based on visual sense Download PDFInfo
- Publication number
- CN103868924A CN103868924A CN201210551389.XA CN201210551389A CN103868924A CN 103868924 A CN103868924 A CN 103868924A CN 201210551389 A CN201210551389 A CN 201210551389A CN 103868924 A CN103868924 A CN 103868924A
- Authority
- CN
- China
- Prior art keywords
- bearing
- image
- detection
- open defect
- real
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Image Analysis (AREA)
Abstract
本发明属于计算机影响领域，是一种对实时轴承外观缺陷检测的方法。本发明首先通过工业相机采集生产线上的轴承图像，然后对图像进行预处理，包括圆心检测，区域分割，图像旋转，阈值分割以及缺陷检测。本发明能够以较快的速度和较高的准确性识别出外观有缺陷的生产线上的轴承。
The invention belongs to the computer-influenced field and relates to a method for detecting real-time bearing appearance defects. In the present invention, the bearing image on the production line is first collected by an industrial camera, and then the image is preprocessed, including circle center detection, region segmentation, image rotation, threshold value segmentation and defect detection. The invention can identify bearings on the production line with defects in appearance at a faster speed and with higher accuracy.
Description
Technical field
The invention belongs to field of computer technology, can detect fast and accurately the defective bearing of outward appearance.The present invention can make bearing surface defect detect operation and realize robotization, make the clearance of defect recognition rate, substandard product reach higher level, improve the oeverall quality of the product that dispatches from the factory, avoid as far as possible actual loss, improve the credit worthiness of enterprise, improve the modernization level of enterprise
Background technology
In Production of bearing process, return and have a small amount of bearing outside surface to have the defect such as corrosion, scuffing, and become useless, substandard products, must these are useless before dispatching from the factory, substandard products identification, reject out.Traditional bearing surface defects detection is mainly to utilize manually to detect, and manual detection is easy to occur flase drop and undetected, not only efficiency low, lack accuracy and standardization, and can not send in real time computing machine and carry out quality management detecting Data classification.In recent years, each manufacturing enterprise is by technological transformation, make to have adopted in Production of bearing robotization unit or automatic production line, and stride forward to modernization, but have surface defects detection only and still adopt artificial visually examine's method with rejecting, not only efficiency is low, and poor reliability, therefore automatic detection instrument is badly in need of in each enterprise, substitutes manual operation.Load mould, scuffing and size thereof, the degree of depth and distributing position on bearing surface is all random.Therefore, with contact measurement, not only difficulty is large, and efficiency is low, adopts vision-based detection to make non-contact detecting, is the preferred approach addressing this problem.Computer Vision Detection Technique is detected for bearing surface defect, and the efficiency that can make up Traditional Man detection is low, and the defects such as fallout ratio height, have the advantages such as efficiency is high, fallout ratio is low, accurate positioning, anti-noise ability is strong, computing velocity is fast, real-time is good.
The present invention can make bearing surface defect detect operation and realize robotization, make the clearance of defect recognition rate, substandard product reach higher level, improve the oeverall quality of the product that dispatches from the factory, avoid as far as possible actual loss, improve the credit worthiness of enterprise, improve the modernization level of enterprise.This project implementation, not only can reduce the loss causing because of the product quality problem that dispatches from the factory, and can greatly cut down the number that detects workman simultaneously.Therefore, enterprise can reduce cost of labor greatly, improves output and efficiency, thereby increases enterprise profit, if implement this project at all relevant enterprises, its economic benefit will be very considerable, and the application prospect of this technology is extremely wide.
Summary of the invention
A kind of method that the object of the present invention is to provide real-time bearing open defect to detect.Specific implementation comprises the following steps:
(1) collection of bearing appearance images: gather image storage by industrial camera acquisition system;
(2) pre-service of bearing appearance images: extract bearing image-region, and it is carried out to the level and smooth pre-service of denoising, and carry out center of circle detection;
(3) extraction of bearing outward appearance: utilize Canny operator edge detection to find out bearing profile;
(4) identification of bearing open defect: utilize the poor method that subtracts coupling to detect the defective bearing of outward appearance.
The invention has the advantages that identification accurately, and processing speed is fast.
Brief description of the drawings
Fig. 1 is the hardware block diagram of the image capturing system that uses in the present invention;
Fig. 2 a, 2b, 2c, 2d, 2e, 2f are the schematic diagram of bearing image processing step.
Embodiment
Below in conjunction with accompanying drawing and instantiation, the present invention will be further described:
A) collection of bearing appearance images:
By reference to the accompanying drawings 1, system, under the irradiation of light-source system, gathers bearing surface image by industrial camera.By image pick-up card, the digitized signal of image is deposited in calculator memory, obtain the gray level image of bearing.In the time that bearing surface exists the defect such as cut, pit, gray level image is after a series of images Processing Algorithm, and its image must have different with the image of standard rolling bearing, utilizes both difference, selects suitable algorithm to divide and detects certified products and non-certified products.Finally carry out automatic classification according to testing result, utilize rejecting mechanism to reject underproof bearing.
B) pre-service of bearing appearance images:
By reference to the accompanying drawings 2, this step method is divided into following a few part:
First step noise-removed filtering.Adopt medium filtering to process original image, result is as accompanying drawing 2a.(a) figure in accompanying drawing 2a represents original image RGB image, and (b) figure represents original image gray level image, and (c) figure represents the image after gray level image medium filtering;
The second step center of circle is detected.According to the actual conditions of the image collecting.In the image collecting, appoint and get a pixel M (must guarantee the inside of M point at bearing inner race), then search out successively four some A1, A2, B1, B2 on bearing inner race left, to the right, upwards, downwards, then get the horizontal ordinate that the horizontal ordinate of A1A2 mid point is ordered as bearing center position O, the ordinate that the ordinate of B1B2 mid point is ordered as bearing center position O.In order to improve accuracy, the position double counting that change M is ordered three times, gets the mean value of two groups of wherein close data, finally determines the coordinate that bearing center O is ordered.Result is as accompanying drawing 2b;
The 3rd step Region Segmentation.According to the size of the size of image and bearing reality, demarcation that just can completion system, thus be partitioned into inner ring, outer ring and have mint-mark word segment.Result is as accompanying drawing 2c.(a) figure in accompanying drawing 2c represents the bearing inner race outer ring image after Region Segmentation, and (b) figure represents the O-ring seal word segment image after Region Segmentation;
The 4th step image rotating.Gland bonnet end face, by centered by the center of circle, is divided into a sector region every 2 °, and the black picture element in sector region is counted and added up.According to the statistical value of black pixel point in sector region, compare with standard form, determine the anglec of rotation of detected bearing image, and image rotating.Can obtain the approximate anglec of rotation of mint-mark word relative standard bearing, and be rotated with this angle.Result is as accompanying drawing 2d.The accurate workpiece image of (a) chart indicating in accompanying drawing 2d, (b) figure in accompanying drawing 2d represents measured workpiece image, image after (c) figure expression measured workpiece image rotation in accompanying drawing 2d;
The 5th step Threshold segmentation.Result is as accompanying drawing 2e.(a) figure in accompanying drawing 2e represents original image, and (b) figure in accompanying drawing 2e represents the image after original image Threshold segmentation;
The 6th step defects detection.Bearing surface is divided into has mint-mark word (gland bonnet part) and without mint-mark word (bearing enclose) two parts, respectively these two parts is carried out subtracting and mating with the difference of standard picture.It should be noted that, before O-ring seal word segment is differed from subtracting coupling, must first be rotated bearing image, and postrotational image is carried out to rim detection, final to the edge detecting is differed from and subtracts coupling, the image of regarding as word in this algorithm is processed.First rotate the image of measured bearing, make it consistent with the word anglec of rotation of standard rolling bearing image, then Threshold segmentation is carried out in word region, and utilize canny operator to carry out rim detection.Result is as accompanying drawing 2f.Image after (a) figure expression standard workpiece threshold segmentation in accompanying drawing 2f, image after (b) figure expression measured workpiece threshold segmentation in accompanying drawing 2f, the image of (c) chart indicating quasi-element outer ring in accompanying drawing 2f after rim detection, (d) figure in accompanying drawing 2f represents the image of detected element outer ring after rim detection, the image of (e) figure presentation video (a) in accompanying drawing 2f after rim detection, the image of (f) figure presentation video (b) in accompanying drawing 2f after rim detection, (g) figure in accompanying drawing 2f represents that outer ring difference subtracts result, (h) figure in accompanying drawing 2f represents the poor result that subtracts of O-ring seal.
Claims (4)
1. the method that real-time bearing open defect detects, is characterized in that, the method contains following steps:
(1) collection of bearing appearance images: gather image storage by industrial camera acquisition system;
(2) pre-service of bearing appearance images: extract bearing image-region, and it is carried out to the level and smooth pre-service of denoising, and carry out center of circle detection;
(3) extraction of bearing outward appearance: utilize Canny operator edge detection to find out bearing profile;
(4) identification of bearing open defect: utilize the poor method that subtracts coupling to detect the defective bearing of outward appearance.
2. bearing open defect detection method according to claim 1, is characterized in that: it is that basis is to be detected and decided by rim detection and the center of circle that the bearing image-region of described step (2) extracts.What denoising adopted is medium filtering.
3. real-time bearing open defect detection method according to claim 1, is characterized in that: first described step (3) utilizing before Canny operator edge detection and will simplify and calculate and weaken noise through arest neighbors image interpolation convergent-divergent.
4. real-time bearing open defect detection method according to claim 1, it is characterized in that: described step (4) utilizes the poor method that subtracts coupling will be through pretreated bearing image and standard rolling bearing image ratio pair, thereby detects the defective bearing of outward appearance.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210551389.XA CN103868924A (en) | 2012-12-18 | 2012-12-18 | Bearing appearance defect detecting algorithm based on visual sense |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210551389.XA CN103868924A (en) | 2012-12-18 | 2012-12-18 | Bearing appearance defect detecting algorithm based on visual sense |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103868924A true CN103868924A (en) | 2014-06-18 |
Family
ID=50907689
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210551389.XA Pending CN103868924A (en) | 2012-12-18 | 2012-12-18 | Bearing appearance defect detecting algorithm based on visual sense |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103868924A (en) |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104966300A (en) * | 2015-06-19 | 2015-10-07 | 浙江理工大学 | Bearing roller image detection system, method and image detection device |
| CN105784721A (en) * | 2016-04-28 | 2016-07-20 | 宁波百加百测控设备有限公司 | Machine vision based bearing detector |
| CN106568783A (en) * | 2016-11-08 | 2017-04-19 | 广东工业大学 | Hardware part defect detecting system and method |
| CN106914428A (en) * | 2017-01-16 | 2017-07-04 | 哈尔滨理工大学 | A kind of New Algorithm of the steel ball surface defect Differential Detection based on machine vision |
| CN107314900A (en) * | 2017-07-13 | 2017-11-03 | 宁波环驰太平洋轴承有限公司 | A kind of full-automatic bearing detection machine |
| CN107607536A (en) * | 2017-09-27 | 2018-01-19 | 镇江金利源轴承有限公司 | A kind of bearing outward appearance compares instrument |
| CN109029993A (en) * | 2018-06-20 | 2018-12-18 | 中国计量大学 | In conjunction with the bearing fault detection algorithm of genetic algorithm optimization parameter and machine vision |
| CN109740507A (en) * | 2018-12-29 | 2019-05-10 | 国网浙江省电力有限公司 | A method for abnormal sensing of substation equipment |
| CN110378902A (en) * | 2019-09-11 | 2019-10-25 | 征图新视(江苏)科技股份有限公司 | A kind of scratch detection method under strong noise background |
| CN111389765A (en) * | 2020-03-31 | 2020-07-10 | 上海电气集团股份有限公司 | Product surface quality detection method and device and product sorting system |
| CN113884299A (en) * | 2021-12-02 | 2022-01-04 | 武汉市书豪塑胶有限公司 | Rotational molding machine fault detection method based on artificial intelligence |
| CN114046729A (en) * | 2021-10-27 | 2022-02-15 | 新兴铸管股份有限公司 | Rapid bearing disassembly detection method |
| CN116523922A (en) * | 2023-07-05 | 2023-08-01 | 上海圣德曼铸造海安有限公司 | Bearing surface defect identification method |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US3598907A (en) * | 1968-05-20 | 1971-08-10 | Emhart Corp | Article inspection by successively televised images |
| JPH04188685A (en) * | 1990-11-19 | 1992-07-07 | Hitachi Electron Eng Co Ltd | How to adjust external mirror type laser oscillator |
| CN102636490A (en) * | 2012-04-12 | 2012-08-15 | 江南大学 | Method for detecting surface defects of dustproof cover of bearing based on machine vision |
-
2012
- 2012-12-18 CN CN201210551389.XA patent/CN103868924A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US3598907A (en) * | 1968-05-20 | 1971-08-10 | Emhart Corp | Article inspection by successively televised images |
| JPH04188685A (en) * | 1990-11-19 | 1992-07-07 | Hitachi Electron Eng Co Ltd | How to adjust external mirror type laser oscillator |
| CN102636490A (en) * | 2012-04-12 | 2012-08-15 | 江南大学 | Method for detecting surface defects of dustproof cover of bearing based on machine vision |
Non-Patent Citations (1)
| Title |
|---|
| 蔡巍巍: "轴承外观表面缺陷检测技术研究", 《中国优秀硕士学位论文全文数据库（电子期刊）信息科技辑》 * |
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104966300B (en) * | 2015-06-19 | 2017-09-12 | 浙江理工大学 | Bearing roller image detecting system and method and image detection device |
| CN104966300A (en) * | 2015-06-19 | 2015-10-07 | 浙江理工大学 | Bearing roller image detection system, method and image detection device |
| CN105784721A (en) * | 2016-04-28 | 2016-07-20 | 宁波百加百测控设备有限公司 | Machine vision based bearing detector |
| CN106568783B (en) * | 2016-11-08 | 2019-12-03 | 广东工业大学 | A hardware part defect detection system and method |
| CN106568783A (en) * | 2016-11-08 | 2017-04-19 | 广东工业大学 | Hardware part defect detecting system and method |
| CN106914428A (en) * | 2017-01-16 | 2017-07-04 | 哈尔滨理工大学 | A kind of New Algorithm of the steel ball surface defect Differential Detection based on machine vision |
| CN107314900A (en) * | 2017-07-13 | 2017-11-03 | 宁波环驰太平洋轴承有限公司 | A kind of full-automatic bearing detection machine |
| CN107607536A (en) * | 2017-09-27 | 2018-01-19 | 镇江金利源轴承有限公司 | A kind of bearing outward appearance compares instrument |
| CN109029993A (en) * | 2018-06-20 | 2018-12-18 | 中国计量大学 | In conjunction with the bearing fault detection algorithm of genetic algorithm optimization parameter and machine vision |
| CN109740507A (en) * | 2018-12-29 | 2019-05-10 | 国网浙江省电力有限公司 | A method for abnormal sensing of substation equipment |
| CN110378902A (en) * | 2019-09-11 | 2019-10-25 | 征图新视(江苏)科技股份有限公司 | A kind of scratch detection method under strong noise background |
| CN111389765A (en) * | 2020-03-31 | 2020-07-10 | 上海电气集团股份有限公司 | Product surface quality detection method and device and product sorting system |
| CN114046729A (en) * | 2021-10-27 | 2022-02-15 | 新兴铸管股份有限公司 | Rapid bearing disassembly detection method |
| CN114046729B (en) * | 2021-10-27 | 2023-07-04 | 新兴铸管股份有限公司 | Bearing quick disassembly detection method |
| CN113884299A (en) * | 2021-12-02 | 2022-01-04 | 武汉市书豪塑胶有限公司 | Rotational molding machine fault detection method based on artificial intelligence |
| CN116523922A (en) * | 2023-07-05 | 2023-08-01 | 上海圣德曼铸造海安有限公司 | Bearing surface defect identification method |
| CN116523922B (en) * | 2023-07-05 | 2023-10-20 | 上海圣德曼铸造海安有限公司 | Bearing surface defect identification method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103868924A (en) | Bearing appearance defect detecting algorithm based on visual sense | |
| CN109141232B (en) | An online detection method for disc castings based on machine vision | |
| CN105069790B (en) | A kind of gear open defect fast image detection method | |
| CN104897062B (en) | A visual measurement method and device for the shape and position deviation of parallel holes on different surfaces of parts | |
| CN109900711A (en) | Workpiece defect detection method based on machine vision | |
| CN106824806B (en) | The detection method of low module plastic gear based on machine vision | |
| CN110766684A (en) | Stator surface defect detection system and detection method based on machine vision | |
| CN107525808A (en) | Blister medication classification and the online visible detection method of defect on a kind of production line | |
| CN117649404A (en) | Medicine packaging box quality detection method and system based on image data analysis | |
| CN103454285A (en) | Transmission chain quality detection system based on machine vision | |
| CN101526484A (en) | Bearing defect detecting technique based on embedded-type machine vision | |
| Li et al. | An online visual measurement method for workpiece dimension based on deep learning | |
| CN110503638B (en) | On-line detection method of spiral glue quality | |
| Hong-Seok et al. | Development of an inspection system for defect detection in pressed parts using laser scanned data | |
| CN110070523B (en) | A kind of foreign matter detection method for bottle bottom | |
| CN111523540A (en) | Deep learning-based metal surface defect detection method | |
| CN108470349A (en) | A kind of milli machine part dimension measurement screening technique under movement background | |
| Liu et al. | An automatic system for bearing surface tiny defect detection based on multi-angle illuminations | |
| CN102621154A (en) | Method and device for automatically detecting cloth defects on line based on improved differential box multi-fractal algorithm | |
| CN112907556A (en) | Automatic measuring method for abrasion loss of rotary cutter based on machine vision | |
| CN105184792B (en) | A kind of saw blade wear extent On-line Measuring Method | |
| CN119228805B (en) | A knitted fabric defect detection method and system based on machine vision | |
| Xu et al. | A knowledge augmented deep learning method for vision-based yarn contour detection | |
| CN115060725A (en) | Double-disc appearance detection method and device | |
| CN105631857B (en) | A kind of scratch detection method and apparatus of optical element surface |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20140618 |