CN108074231A - Magnetic sheet surface defect detection method based on convolutional neural network - Google Patents
Magnetic sheet surface defect detection method based on convolutional neural network Download PDFInfo
- Publication number
- CN108074231A CN108074231A CN201711361710.7A CN201711361710A CN108074231A CN 108074231 A CN108074231 A CN 108074231A CN 201711361710 A CN201711361710 A CN 201711361710A CN 108074231 A CN108074231 A CN 108074231A
- Authority
- CN
- China
- Prior art keywords
- image
- magnetic sheet
- layers
- input
- convolutional neural
- 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
Links
- 238000013527 convolutional neural network Methods 0.000 title claims abstract description 36
- 230000007547 defect Effects 0.000 title claims abstract description 32
- 238000001514 detection method Methods 0.000 title claims abstract description 29
- 238000000034 method Methods 0.000 claims abstract description 26
- 238000005070 sampling Methods 0.000 claims abstract description 8
- 238000012545 processing Methods 0.000 claims description 22
- 238000012549 training Methods 0.000 claims description 18
- 239000011159 matrix material Substances 0.000 claims description 15
- 230000001429 stepping effect Effects 0.000 claims description 15
- 230000002950 deficient Effects 0.000 claims description 12
- 238000000605 extraction Methods 0.000 claims description 10
- 230000005284 excitation Effects 0.000 claims description 9
- 239000000284 extract Substances 0.000 claims description 9
- 238000012360 testing method Methods 0.000 claims description 6
- 230000011218 segmentation Effects 0.000 claims description 5
- 201000001371 inclusion conjunctivitis Diseases 0.000 claims description 4
- 206010044325 trachoma Diseases 0.000 claims description 4
- 208000037656 Respiratory Sounds Diseases 0.000 claims description 3
- 230000000694 effects Effects 0.000 claims description 2
- 238000011156 evaluation Methods 0.000 claims description 2
- 239000012141 concentrate Substances 0.000 claims 1
- 238000007781 pre-processing Methods 0.000 abstract description 3
- 238000005520 cutting process Methods 0.000 abstract description 2
- 230000009466 transformation Effects 0.000 abstract 2
- 238000005516 engineering process Methods 0.000 description 7
- 230000007935 neutral effect Effects 0.000 description 3
- 238000007689 inspection Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000003672 processing method Methods 0.000 description 2
- 238000009825 accumulation Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 210000003739 neck Anatomy 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/048—Activation functions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Biophysics (AREA)
- Evolutionary Computation (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Software Systems (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
- Investigating Or Analyzing Materials By The Use Of Magnetic Means (AREA)
Abstract
The invention discloses a magnetic sheet surface defect detection method based on a convolutional neural network, which comprises the following steps: the method comprises the steps of firstly, obtaining an overlook image of a magnetic sheet to be detected and preprocessing the image, wherein the steps comprise graying, Hough circle transformation, size transformation, rotary cutting and the like; secondly, inputting the preprocessed image into a pre-trained convolutional neural network for defect detection, detecting whether the surface of the magnetic sheet has defects, and classifying the defects; the convolutional neural network comprises an input layer, a convolutional layer, a sampling layer and a full-connection layer, wherein the input layer, the convolutional layer and the sampling layer are used for extracting the characteristics of the image, and the extracted characteristics are subjected to defect classification by a Softmax classifier. Compared with the prior art, the method has high detection precision and better robustness.
Description
Technical field
The invention belongs to surface defects detection technology more particularly to a kind of magnetic sheet surface defects based on convolutional neural networks
Detection method.
Background technology
With the rapid development of electronic technology, computer technology, digital image processing techniques are with its information content is big, performance
The advantages that formal intuition, transmission convenient storage, is widely applied in many industries and field, if Medical Image Processing is analyzed,
Industry Control and detection automation, airborne and spaceborne RS mapping etc..With the promotion of Chinese national economy level, people are to height
Quality, high-precision, the product demand of high reliability also increasingly increase, accompanying problem is that the product largely manufactured is such as
What detects, judges whether it reaches performance indicator.
Traditional detection method is detected by manually, and labor strength is big in artificial detection, is limited to worker's
The state of mind, detection proficient, experience accumulation be horizontal and many factors such as working environment, and the efficiency of detection is low, speed
Slowly, the uniformity mark of device is difficult to be guaranteed.Due to worker's fatigue in detection process, inevitable generation mistake is picked up, missing inspection,
The outflow of substandard product can not only bring economic loss to factory, more seriously can bring security risk to user.Therefore,
How rapidly and efficiently and accurately detection surface defects of parts becomes manufacturing industry urgent problem.The promotion of computer technology,
So that digital image processing techniques are used widely, the surface defects detection system based on machine vision has very high application
Value.
There are mainly two types of the methods for being currently based on the surface defects detection of machine vision：One kind is based on image processing algorithm
Characteristic processing method, another kind is the method that is combined using convolutional neural networks with image processing algorithm.At image
The characteristic processing method of adjustment method, be calculated respectively by handling the gray processing of image image average gray numerical value and
The variance of gray processing numerical value, by compared with pre-set standard value, judging whether obtain size of the difference with standard value
More than threshold value to determine whether there is surface defect.This method usually requires pre-set camera exposure time, ambient light
The information such as bright dark, magnetic sheet acquisition position, are affected, the robustness of system is poor, and is asked the defects of solution by external condition
It inscribes single.And the method being combined using convolutional neural networks with image processing algorithm, convolutional network are to identify two-dimensional shapes
And a multilayer perceptron of special designing, this network structure is to translation, proportional zoom, inclination or the deformation of his form altogether
With height consistency.And avoid feature extraction complicated in tional identification algorithm and data reconstruction processes.Defect characteristic
Recognition speed can also improve a lot, but such method needs substantial amounts of sample to train the discrimination for having reached higher,
Training sample is very few discrimination to be caused to decline.
Surface detection technique of the present invention relate generally to Image Acquisition, image processing, convolutional neural networks structure and
Trained, image deflects classification and etc..Wherein the digital processing of image and the structure of convolutional neural networks are Surface testing necks
Two key issues in domain.
The content of the invention
The present invention will overcome the disadvantages mentioned above of the prior art, provide a kind of magnetic sheet surface defect based on convolutional neural networks
Detection method, so as to fast and accurately realize the detection of magnetic sheet surface number of drawbacks.
In order to achieve the above objectives, technical scheme is specifically realized in：
A kind of magnetic sheet detection method of surface flaw based on convolutional neural networks, this method comprises the following steps：
(1) obtain the overhead view image of magnetic sheet to be detected and image is pre-processed；
(2) pretreated image is inputted to trained convolutional neural networks in advance and carries out defects detection, detection magnetic
Whether piece surface is defective, and classifies to defect；Convolutional neural networks, wherein input layer, convolutional layer, sample level, Quan Lian
It connects layer and feature extraction is carried out to image, the feature of extraction carries out defect classification by Softmax graders.
The pretreatment of magnetic sheet image specifically comprises the following steps：
The first step carries out gray processing processing to magnetic sheet image, carries out hough-circle transform to gray level image, detects outside magnetic sheet
Profile, according to the center of circle, radius of circle cuts the external square of minimum of the circle；
Second step is the batch template matching treatment residual image so that institute using the square-shaped image being cut into as template
There is the outer profile minimum circumscribed circle size that image is magnetic sheet；
All images that second step obtains are carried out size change over to being sized by the 3rd step；
Image after size change over is carried out rotation segmentation by the 4th step using the method for image procossing, by a secondary big figure segmentation
Into several small figures；
The training method of convolutional neural networks is as follows：
(a) data sample gathers.A large amount of magnetic sheet images are gathered, include zero defect magnetic sheet and defective magnetic sheet；
(b) EDS extended data set.The magnetic sheet of acquisition is pre-processed, image carries out rotation segmentation, and a secondary magnetic sheet figure is cut
Multiple image is cut into, and carries out handmarking, is divided into defective trachoma class, defective scarce block class, defective crackle class, zero defect
Totally 4 class, and these data samples to obtaining are divided into two parts by a certain percentage, for a part as training set, a part is survey
Examination collection；
(c) convolutional neural networks are established；
(d) training set sample data is inputted to convolutional neural networks, training convolutional neural networks, and with test set sample
Data go the effect of evaluation convolutional neural networks training；
The structure of convolutional neural networks is as follows：
Input layer：First layer is input layer, and input size is picture size 43*43；
C1 layers：The output of input layer is the matrix of 43*43, and the input of C1 convolutional layers is the output of input layer, is inputted and 64
The convolution kernel that a 3*3 steppings are 1 carries out convolution operation, and excitation function RELU extracts 64 feature map, exports as 41* in total
The matrix of 41*64；
S2 layers：The input of S2 ponds layer is the output of C1 convolutional layers, to 64 map convolution that 2*2 steppings are 2 of input
Core carries out maximum sampling processing, exports the matrix for 20*20*64；
C3 layers：C3 layers of input is S2 layers of output, to 64 map convolution kernels that 128 3*3 steppings are 1 of input
Examination paper operation is carried out, excitation function RELU extracts 128 feature map, exports the matrix for 18*18*128 in total；
S4 layers：S4 layers of input is C3 layers of output, and 128 map of input are carried out with the convolution kernel that 2*2 steppings are 2
Maximum sampling processing exports the matrix for 9*9*128；
C5 layers：C5 layers of input is S4 layers of output, to 128 map convolution kernels that 256 3*3 steppings are 1 of input
Convolution operation is carried out, excitation function RELU extracts 256 feature map, exports as 7*7*256 in total；
The full articulamentums of F6：The feature of extraction is weighted in full articulamentum 1, obtains 1 dimensional feature vector；
The full articulamentums 2 of F7：The feature vector that full articulamentum 2 exports full articulamentum 1 is weighted, and obtains feature more
1 dimensional feature vector concentrated；
Softmax graders：Classify to the feature vector of complete 2 output terminal of articulamentum；
Compared with prior art, the present invention its remarkable advantage：(1) pretreatment of the present invention to image to be detected, is equivalent to
Training samples number is added, improves the problem of neural metwork training caused by training sample lacks is bad.(2) traditional base
The detection method in the digital picture the defects of, recognizer contain complicated feature extraction and data reconstruction processes, and each defect
Individual recognizer is required for, but the present invention can effectively avoid problem above, trained volume using convolutional neural networks
Product neutral net can classify number of drawbacks.(3) using convolutional neural networks compared to traditional detection based on image procossing
Method robustness higher, in image acquisition phase, image is influenced be subject to extraneous factor, such as the variation of light, conventional method have
May extraction less than feature, but the method for convolutional neural networks can be purposefully added when training neutral net part by
The picture that extraneous factor influences strengthens Detection accuracy of the convolutional neural networks for this respect, this point more meets industry spot
Requirement.
Description of the drawings
Fig. 1 is the flow chart of the method for the present invention；
Fig. 2 is present invention magnetic sheet surface image preprocessing flow chart to be detected；
Fig. 3 is the pretreated magnetic sheet surface image to be detected of implementation；
Fig. 4 is the structure diagram of the network of convolutional neural networks of the present invention；
Specific embodiment
With reference to the attached drawing in the embodiment of the present invention, the technical solution in the embodiment of the present invention is carried out clear, complete
Ground describes.
With reference to Fig. 1~Fig. 4, a kind of magnetic sheet defect inspection method based on convolutional neural networks, including following process：First
Training material is collected, there are 1000 containing defective magnetic sheet and 500 zero defect magnetic sheets.The two of magnetic sheet is gathered under standard environment
Face top view by image preprocessing process shown in Fig. 2, first carries out gray processing processing to magnetic sheet image, gray level image is carried out
Hough-circle transform detects magnetic sheet outer profile, and according to the center of circle, radius of circle cuts the external square of minimum of the circle；Again with cutting
Square-shaped image out be template, batch template matching treatment residual image so that all images are the outer profile of magnetic sheet
Minimum circumscribed circle size carries out all images size change over to being sized, rotation point is carried out using the method for image procossing
It cuts, a secondary big figure is divided into several small figures.Gained image is manually demarcated, and after balancing each defective proportion, is obtained
1000 trachoma classes, 1000 scarce block classes, 1000 cut classes and 1000 zero defect picture materials, wherein each selection
700 are put into training set, and 300 are put into test set.Training set and test set are sequentially placed into the convolution under Tensorflow frameworks
Neutral net, trained convolutional neural networks are just the model of detection magnetic sheet defect.
The structure of convolutional neural networks is as follows：
Input layer：First layer is input layer, and input size is picture size 43*43；
C1 layers：The output of input layer is the matrix of 43*43, and the input of C1 convolutional layers is the output of input layer, is inputted and 64
The convolution kernel that a 3*3 steppings are 1 carries out convolution operation, and excitation function RELU extracts 64 feature map, exports as 41* in total
The matrix of 41*64；
S2 layers：The input of S2 ponds layer is the output of C1 convolutional layers, to 64 map convolution that 2*2 steppings are 2 of input
Core carries out maximum sampling processing, exports the matrix for 20*20*64；
C3 layers：C3 layers of input is S2 layers of output, to 64 map convolution kernels that 128 3*3 steppings are 1 of input
Examination paper operation is carried out, excitation function RELU extracts 128 feature map, exports the matrix for 18*18*128 in total；
S4 layers：S4 layers of input is C3 layers of output, and 128 map of input are carried out with the convolution kernel that 2*2 steppings are 2
Maximum sampling processing exports the matrix for 9*9*128；
C5 layers：C5 layers of input is S4 layers of output, to 128 map convolution kernels that 256 3*3 steppings are 1 of input
Convolution operation is carried out, excitation function RELU extracts 256 feature map, exports as 7*7*256 in total；
The full articulamentums of F6：The feature of extraction is weighted in full articulamentum 1, obtains 1 dimensional feature vector；
The full articulamentums 2 of F7：The feature vector that full articulamentum 2 exports full articulamentum 1 is weighted, and obtains feature more
1 dimensional feature vector concentrated；
Softmax graders：Classify to the feature vector of complete 2 output terminal of articulamentum；
Output layer includes 5 units, represents 5 kinds of situations respectively：(1) it is intact to lack block class (3) trachoma class (4) for crackle class (2)
Fall into class (5) other classes.
Content described in this specification embodiment is only enumerating to the way of realization of inventive concept, protection of the invention
Scope is not construed as being only limitted to the concrete form that embodiment is stated, protection scope of the present invention is also and in art technology
Personnel conceive according to the present invention it is conceivable that equivalent technologies mean.
Claims (1)
1. a kind of magnetic sheet detection method of surface flaw based on convolutional neural networks, includes the following steps：
(1) obtain the overhead view image of magnetic sheet to be detected and image is pre-processed；
Specifically comprise the following steps：
The first step carries out gray processing processing to magnetic sheet image, carries out hough-circle transform to gray level image, detects magnetic sheet outer profile,
According to the center of circle, radius of circle cuts the external square of minimum of the circle；
Second step is the batch template matching treatment residual image so that Suo Youtu using the square-shaped image being cut into as template
Outer profile minimum circumscribed circle size as being magnetic sheet；
All images that second step obtains are carried out size change over to being sized by the 3rd step；
Image after size change over is carried out rotation segmentation by the 4th step using the method for image procossing, a secondary big figure is divided into more
Small figure；
(2) pretreated image is inputted to trained convolutional neural networks in advance and carries out defects detection, detect magnetic disk sheet
Whether face is defective, and classifies to defect；Convolutional neural networks, wherein input layer, convolutional layer, sample level, full articulamentum
Feature extraction is carried out to image, the feature of extraction carries out defect classification by Softmax graders；
The defects of described, the training method of detection model was：
(a) data sample gathers, and gathers a large amount of magnetic sheet images, includes zero defect magnetic sheet and defective magnetic sheet；
(b) EDS extended data set pre-processes the magnetic sheet of acquisition, and image carries out rotation segmentation, and a secondary magnetic sheet figure is cut into
Multiple image, and carry out handmarking, is divided into defective trachoma class, defective scarce block class, defective crackle class, zero defect totally 4
Class, and these data samples to obtaining are divided into two parts by a certain percentage, for a part as training set, a part is test
Collection；
(c) convolutional neural networks are established；
(d) training set sample data is inputted to convolutional neural networks, training convolutional neural networks, and with test set sample data
Go the effect of evaluation convolutional neural networks training；
The convolutional neural networks are as follows：
Input layer：First layer is input layer, and input size is picture size 43*43；
C1 layers：The output of input layer is the matrix of 43*43, and the input of C1 convolutional layers is the output of input layer, is inputted and 64 3*3
The convolution kernel that stepping is 1 carries out convolution operation, and excitation function RELU extracts 64 feature map, exports as 41*41*64 in total
Matrix；
S2 layers：The input of S2 ponds layer for C1 convolutional layers output, to 64 map of input with the convolution kernel that 2*2 steppings are 2 into
Row maximum sampling processing, exports the matrix for 20*20*64；
C3 layers：C3 layers of input is S2 layers of output, and 64 map of input are carried out with the convolution kernel that 128 3*3 steppings are 1
Examination paper operates, and excitation function RELU extracts 128 feature map, exports the matrix for 18*18*128 in total；
S4 layers：S4 layers of input is C3 layers of output, and maximum is carried out with the convolution kernel that 2*2 steppings are 2 to 128 map of input
It is worth sampling processing, exports the matrix for 9*9*128；
C5 layers：C5 layers of input is S4 layers of output, and 128 map of input are carried out with the convolution kernel that 256 3*3 steppings are 1
Convolution operation, excitation function RELU extract 256 feature map, export as 7*7*256 in total；
The full articulamentums of F6：The feature of extraction is weighted in full articulamentum 1, obtains 1 dimensional feature vector；
The full articulamentums 2 of F7：The feature vector that full articulamentum 2 exports full articulamentum 1 is weighted, and obtains feature and more concentrates
1 dimensional feature vector；
Softmax graders：Classify to the feature vector of complete 2 output terminal of articulamentum.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711361710.7A CN108074231B (en) | 2017-12-18 | 2017-12-18 | Magnetic sheet surface defect detection method based on convolutional neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711361710.7A CN108074231B (en) | 2017-12-18 | 2017-12-18 | Magnetic sheet surface defect detection method based on convolutional neural network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108074231A true CN108074231A (en) | 2018-05-25 |
| CN108074231B CN108074231B (en) | 2020-04-21 |
Family
ID=62158463
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711361710.7A Active CN108074231B (en) | 2017-12-18 | 2017-12-18 | Magnetic sheet surface defect detection method based on convolutional neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108074231B (en) |
Cited By (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108685158A (en) * | 2018-05-28 | 2018-10-23 | 杭州安脉盛智能技术有限公司 | Filament cutter intelligent maintenance method and system based on automatic optics inspection technology |
| CN108960413A (en) * | 2018-07-11 | 2018-12-07 | 天津工业大学 | A kind of depth convolutional neural networks method applied to screw surface defects detection |
| CN108985337A (en) * | 2018-06-20 | 2018-12-11 | 中科院广州电子技术有限公司 | A kind of product surface scratch detection method based on picture depth study |
| CN109064454A (en) * | 2018-07-12 | 2018-12-21 | 上海蝶鱼智能科技有限公司 | Product defects detection method and system |
| CN109239075A (en) * | 2018-08-27 | 2019-01-18 | 北京百度网讯科技有限公司 | Battery detection method and device |
| CN109300109A (en) * | 2018-08-23 | 2019-02-01 | 苏州富鑫林光电科技有限公司 | The industrial intelligent image segmentation and identifying system and method for machine learning |
| CN109376773A (en) * | 2018-09-30 | 2019-02-22 | 福州大学 | Crack detecting method based on deep learning |
| CN109685968A (en) * | 2018-12-15 | 2019-04-26 | 西安建筑科技大学 | A kind of the identification model building and recognition methods of the banknote image defect based on convolutional neural networks |
| CN109800708A (en) * | 2018-12-13 | 2019-05-24 | 程琳 | Visit image lesion intelligent identification Method in aero-engine hole based on deep learning |
| CN110243934A (en) * | 2019-05-30 | 2019-09-17 | 中国计量大学上虞高等研究院有限公司 | A kind of ultrasonic weld seam detection method based on wavelet convolution neural network |
| CN110288013A (en) * | 2019-06-20 | 2019-09-27 | 杭州电子科技大学 | A kind of defective labels recognition methods based on block segmentation and the multiple twin convolutional neural networks of input |
| CN110929773A (en) * | 2019-11-18 | 2020-03-27 | 上海集成电路研发中心有限公司 | Chip surface defect classification device and method based on deep learning |
| CN111062391A (en) * | 2019-12-25 | 2020-04-24 | 创新奇智(青岛)科技有限公司 | Initial positioning method for magnetic sheet |
| CN111292345A (en) * | 2019-05-17 | 2020-06-16 | 展讯通信（上海）有限公司 | Purple border detection method and device, electronic equipment and storage medium |
| CN112419291A (en) * | 2020-11-30 | 2021-02-26 | 佛山职业技术学院 | Training method of bottle blank defect detection model, storage medium and terminal equipment |
| CN112819824A (en) * | 2021-03-03 | 2021-05-18 | 科为升视觉技术（苏州）有限公司 | Neural network of visual inspection system, and inspection system and method including the same |
| CN113516650A (en) * | 2021-07-30 | 2021-10-19 | 深圳康微视觉技术有限公司 | Circuit board hole plugging defect detection method and device based on deep learning |
| CN113610795A (en) * | 2021-08-02 | 2021-11-05 | 沈阳航空航天大学 | Combustible cartridge case surface defect detection method and system |
| CN113793332A (en) * | 2021-11-15 | 2021-12-14 | 山东德普检测技术有限公司 | Experimental instrument defect identification and classification method and system |
| CN116091505A (en) * | 2023-04-11 | 2023-05-09 | 青岛芯康半导体科技有限公司 | Automatic defect detection and classification method and system for sapphire substrate |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106127780A (en) * | 2016-06-28 | 2016-11-16 | 华南理工大学 | A kind of curved surface defect automatic testing method and device thereof |
| US20170140524A1 (en) * | 2015-11-17 | 2017-05-18 | Kla-Tencor Corporation | Single image detection |
| CN107169556A (en) * | 2017-05-15 | 2017-09-15 | 电子科技大学 | stem cell automatic counting method based on deep learning |
| CN107316295A (en) * | 2017-07-02 | 2017-11-03 | 苏州大学 | A kind of fabric defects detection method based on deep neural network |
| CN107392896A (en) * | 2017-07-14 | 2017-11-24 | 佛山市南海区广工大数控装备协同创新研究院 | A kind of Wood Defects Testing method and system based on deep learning |
| CN107481231A (en) * | 2017-08-17 | 2017-12-15 | 广东工业大学 | A kind of handware defect classifying identification method based on depth convolutional neural networks |
-
2017
- 2017-12-18 CN CN201711361710.7A patent/CN108074231B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20170140524A1 (en) * | 2015-11-17 | 2017-05-18 | Kla-Tencor Corporation | Single image detection |
| CN106127780A (en) * | 2016-06-28 | 2016-11-16 | 华南理工大学 | A kind of curved surface defect automatic testing method and device thereof |
| CN107169556A (en) * | 2017-05-15 | 2017-09-15 | 电子科技大学 | stem cell automatic counting method based on deep learning |
| CN107316295A (en) * | 2017-07-02 | 2017-11-03 | 苏州大学 | A kind of fabric defects detection method based on deep neural network |
| CN107392896A (en) * | 2017-07-14 | 2017-11-24 | 佛山市南海区广工大数控装备协同创新研究院 | A kind of Wood Defects Testing method and system based on deep learning |
| CN107481231A (en) * | 2017-08-17 | 2017-12-15 | 广东工业大学 | A kind of handware defect classifying identification method based on depth convolutional neural networks |
Non-Patent Citations (1)
| Title |
|---|
| R. LIU ETC: "Regionconvolutional neural network for detecting capsule surface defects", 《BOLETIN TECNICO/TECHNICAL BULLETIN》 * |
Cited By (28)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108685158A (en) * | 2018-05-28 | 2018-10-23 | 杭州安脉盛智能技术有限公司 | Filament cutter intelligent maintenance method and system based on automatic optics inspection technology |
| CN108685158B (en) * | 2018-05-28 | 2021-11-09 | 杭州安脉盛智能技术有限公司 | Intelligent filament cutter maintenance method and system based on automatic optical detection technology |
| CN108985337A (en) * | 2018-06-20 | 2018-12-11 | 中科院广州电子技术有限公司 | A kind of product surface scratch detection method based on picture depth study |
| CN108960413A (en) * | 2018-07-11 | 2018-12-07 | 天津工业大学 | A kind of depth convolutional neural networks method applied to screw surface defects detection |
| CN109064454A (en) * | 2018-07-12 | 2018-12-21 | 上海蝶鱼智能科技有限公司 | Product defects detection method and system |
| CN109300109A (en) * | 2018-08-23 | 2019-02-01 | 苏州富鑫林光电科技有限公司 | The industrial intelligent image segmentation and identifying system and method for machine learning |
| CN109300109B (en) * | 2018-08-23 | 2020-10-09 | 苏州富鑫林光电科技有限公司 | Machine learning industrial intelligent image segmentation and recognition system |
| CN109239075A (en) * | 2018-08-27 | 2019-01-18 | 北京百度网讯科技有限公司 | Battery detection method and device |
| CN109376773A (en) * | 2018-09-30 | 2019-02-22 | 福州大学 | Crack detecting method based on deep learning |
| WO2020119103A1 (en) * | 2018-12-13 | 2020-06-18 | 程琳 | Aero-engine hole detection image damage intelligent identification method based on deep learning |
| CN109800708A (en) * | 2018-12-13 | 2019-05-24 | 程琳 | Visit image lesion intelligent identification Method in aero-engine hole based on deep learning |
| CN109685968A (en) * | 2018-12-15 | 2019-04-26 | 西安建筑科技大学 | A kind of the identification model building and recognition methods of the banknote image defect based on convolutional neural networks |
| CN111292345A (en) * | 2019-05-17 | 2020-06-16 | 展讯通信（上海）有限公司 | Purple border detection method and device, electronic equipment and storage medium |
| CN110243934A (en) * | 2019-05-30 | 2019-09-17 | 中国计量大学上虞高等研究院有限公司 | A kind of ultrasonic weld seam detection method based on wavelet convolution neural network |
| CN110288013A (en) * | 2019-06-20 | 2019-09-27 | 杭州电子科技大学 | A kind of defective labels recognition methods based on block segmentation and the multiple twin convolutional neural networks of input |
| CN110929773A (en) * | 2019-11-18 | 2020-03-27 | 上海集成电路研发中心有限公司 | Chip surface defect classification device and method based on deep learning |
| CN111062391B (en) * | 2019-12-25 | 2023-09-19 | 创新奇智(青岛)科技有限公司 | Magnetic sheet initial positioning method |
| CN111062391A (en) * | 2019-12-25 | 2020-04-24 | 创新奇智(青岛)科技有限公司 | Initial positioning method for magnetic sheet |
| CN112419291A (en) * | 2020-11-30 | 2021-02-26 | 佛山职业技术学院 | Training method of bottle blank defect detection model, storage medium and terminal equipment |
| CN112419291B (en) * | 2020-11-30 | 2024-06-04 | 佛山职业技术学院 | Training method of bottle embryo defect detection model, storage medium and terminal equipment |
| CN112819824A (en) * | 2021-03-03 | 2021-05-18 | 科为升视觉技术（苏州）有限公司 | Neural network of visual inspection system, and inspection system and method including the same |
| CN112819824B (en) * | 2021-03-03 | 2024-04-26 | 科为升视觉技术（苏州）有限公司 | Neural network for vision inspection system, inspection system and method including the same |
| CN113516650A (en) * | 2021-07-30 | 2021-10-19 | 深圳康微视觉技术有限公司 | Circuit board hole plugging defect detection method and device based on deep learning |
| CN113516650B (en) * | 2021-07-30 | 2023-08-25 | 深圳康微视觉技术有限公司 | Circuit board hole plugging defect detection method and device based on deep learning |
| CN113610795A (en) * | 2021-08-02 | 2021-11-05 | 沈阳航空航天大学 | Combustible cartridge case surface defect detection method and system |
| CN113610795B (en) * | 2021-08-02 | 2023-09-29 | 沈阳航空航天大学 | Method and system for detecting surface defects of combustible cartridge |
| CN113793332A (en) * | 2021-11-15 | 2021-12-14 | 山东德普检测技术有限公司 | Experimental instrument defect identification and classification method and system |
| CN116091505A (en) * | 2023-04-11 | 2023-05-09 | 青岛芯康半导体科技有限公司 | Automatic defect detection and classification method and system for sapphire substrate |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108074231B (en) | 2020-04-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108074231A (en) | Magnetic sheet surface defect detection method based on convolutional neural network | |
| CN107316300B (en) | Tire X-ray defect detection method based on deep convolutional neural network | |
| CN111402226A (en) | Surface defect detection method based on cascade convolution neural network | |
| CN108921201B (en) | Dam defect identification and classification method based on feature combination and CNN | |
| CN108918527A (en) | A kind of printed matter defect inspection method based on deep learning | |
| CN109583324A (en) | A kind of pointer meters reading automatic identifying method based on the more box detectors of single-point | |
| CN106355579A (en) | Defect detecting method of cigarette carton surface wrinkles | |
| CN109583295B (en) | Automatic detection method for switch machine notch based on convolutional neural network | |
| CN108932712A (en) | A kind of rotor windings quality detecting system and method | |
| CN108711148A (en) | A kind of wheel tyre defect intelligent detecting method based on deep learning | |
| CN113869162A (en) | Violation identification method and system based on artificial intelligence | |
| CN105930852A (en) | Method for identifying bubble image | |
| CN111368825A (en) | Pointer positioning method based on semantic segmentation | |
| CN115631472B (en) | Intelligent detection method for pedestrian intrusion on expressway | |
| CN113191374A (en) | PolSAR image ridge line extraction method based on pyramid attention network | |
| CN111242899A (en) | Image-based flaw detection method and computer-readable storage medium | |
| CN103886332A (en) | Method for detecting and identifying defects of metallic meshes | |
| CN111291818B (en) | Non-uniform class sample equalization method for cloud mask | |
| CN112560896A (en) | Fruit quality screening and classifying system based on image processing | |
| CN114841957A (en) | Steel plate surface defect detection method based on deep learning | |
| CN113128555B (en) | Method for detecting abnormality of train brake pad part | |
| CN117237747A (en) | Hardware defect classification and identification method based on artificial intelligence | |
| CN112396580A (en) | Circular part defect detection method | |
| CN116091818B (en) | Pointer type instrument reading identification method based on multi-neural network cascading model | |
| CN114119562B (en) | Brake disc outer surface defect detection method and system based on deep learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |