CN111768363A - Deep learning-based circuit board surface defect detection method and detection system - Google Patents
Deep learning-based circuit board surface defect detection method and detection system Download PDFInfo
- Publication number
- CN111768363A CN111768363A CN202010401659.3A CN202010401659A CN111768363A CN 111768363 A CN111768363 A CN 111768363A CN 202010401659 A CN202010401659 A CN 202010401659A CN 111768363 A CN111768363 A CN 111768363A
- Authority
- CN
- China
- Prior art keywords
- circuit board
- image
- defect
- detection
- deep learning
- 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
- 230000007547 defect Effects 0.000 title claims abstract description 131
- 238000001514 detection method Methods 0.000 title claims abstract description 88
- 238000013135 deep learning Methods 0.000 title claims abstract description 26
- 238000005520 cutting process Methods 0.000 claims abstract description 6
- 238000000034 method Methods 0.000 claims description 15
- 238000007781 pre-processing Methods 0.000 claims description 13
- 239000011159 matrix material Substances 0.000 claims description 12
- 238000005286 illumination Methods 0.000 claims description 9
- 238000007689 inspection Methods 0.000 claims description 6
- 230000008439 repair process Effects 0.000 claims description 5
- 238000013136 deep learning model Methods 0.000 claims description 4
- 238000002372 labelling Methods 0.000 claims description 4
- 238000010606 normalization Methods 0.000 claims description 4
- 238000003860 storage Methods 0.000 claims description 4
- 238000010276 construction Methods 0.000 claims description 3
- 239000003550 marker Substances 0.000 claims description 3
- 238000004519 manufacturing process Methods 0.000 description 8
- 230000002950 deficient Effects 0.000 description 4
- 239000000047 product Substances 0.000 description 4
- 238000010586 diagram Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 230000002411 adverse Effects 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 239000012535 impurity Substances 0.000 description 1
- 239000013067 intermediate product Substances 0.000 description 1
- 239000010410 layer Substances 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 239000011148 porous material Substances 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000006748 scratching Methods 0.000 description 1
- 239000002356 single layer Substances 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
Images
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
- G06N3/08—Learning methods
-
- 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]
-
- 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
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30141—Printed circuit board [PCB]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a circuit board surface defect detection method and a detection system based on deep learning, wherein the detection method comprises the following steps: obtaining a detection model of the surface defects of the circuit board; acquiring a surface image of a circuit board to be detected; cutting and normalizing the surface image; taking the preprocessed surface image as an input image of a detection model, detecting the input image by the detection model, adjusting the detected image into an output image of the original circuit board in size, and simultaneously outputting coordinate information, defect types and defect quantity of defect positions in the output image; and controlling the mechanical arm to mark the defect position. The detection method can effectively improve the surface defect detection efficiency of the multi-variety circuit boards.
Description
Technical Field
The invention relates to the technical field of intelligent manufacturing, in particular to a circuit board surface defect detection method and system based on deep learning.
Background
Printed Circuit boards (Circuit boards) are important components essential for electronic products, and realize electrical connection between electronic components. In the production process, the surface of the circuit board often has defects such as open circuit, dirt, scratch and the like due to equipment faults or human factors in the production process, and the surface defects bring adverse effects on the attractiveness, comfort, usability and the like of the circuit board, so that a production enterprise needs to detect the surface defects of the circuit board so as to find defective products in time and effectively control the quality of the defective products.
At present, the defect detection process of domestic circuit board production enterprises is mainly completed by combining AOI (automated optical inspection) and manual detection. Based on AOI detection, the surface defects of the circuit board are generally detected by adopting modes of local pattern matching, image sketch and geometric image comparison and the like aiming at the semi-finished circuit board, the positioning requirements on the circuit board to be detected are very high, and the surface defects of the circuit board with multiple varieties and sizes are difficult to detect; the manual detection is to take pictures of the two sides of each circuit board (including each laminated intermediate product), and scan and detect the circuit boards by human eyes after the circuit boards are amplified in a computer.
With the increasing requirements of electronic products on circuit boards, the circuit boards are developed from single-layer to double-sided and multi-layer, and are continuously developed towards the directions of high precision, high density, fine pore diameter, fine wires and fine spacing, and the size is continuously reduced, so that the difficulty of detecting the surface defects of the circuit boards based on AOI detection and manual detection is gradually increased, and the method is difficult to adapt to the requirement of quickly improving the production efficiency. Therefore, the scheme for rapidly and nondestructively detecting the surface defects of the circuit board is developed, and the method has important significance for effectively controlling the quality of the circuit board, saving the labor cost and improving the production efficiency.
Disclosure of Invention
Aiming at the technical problems in the prior art, the invention aims to: the circuit board surface defect detection method based on deep learning is provided, and the detection efficiency of the surface defects of various circuit boards can be effectively improved.
In order to achieve the above object, an embodiment of a first aspect of the present invention provides a method for detecting surface defects of a circuit board based on deep learning, including the following steps:
obtaining a detection model of the surface defects of the circuit board;
acquiring a surface image of a circuit board to be detected;
preprocessing the surface image;
taking the preprocessed surface image as an input image of a detection model, detecting the input image by the detection model, adjusting the detected image into an output image of the original circuit board in size, and simultaneously outputting coordinate information, defect types and defect quantity of defect positions in the output image;
and controlling the mechanical arm to mark the defect position.
Further, the step of obtaining the detection model of the surface defects of the circuit board comprises the following steps:
acquiring an original image of a surface defect of a standard circuit board;
carrying out random rotation, cutting and normalization pretreatment on an original image;
labeling the preprocessed image by using labelImg and establishing a circuit board defect image data set;
building an improved Faster R-CNN target detection model based on Tensorflow;
a target detection model is trained on the circuit board defect image dataset data.
Further, the circuit board defect image data set comprises defect images of broken lines, sundries, scratches, dirt, poor repaired lines, hole loss, line defect, overproof oil repair and the like on the surface of the circuit board.
Further, pre-processing the surface image includes cropping and normalizing the original image based on a Python programming algorithm.
The embodiment of the second aspect of the invention provides a detection system of a circuit board surface defect detection method based on deep learning, which comprises the following steps:
a model construction unit: the detection model is used for constructing the surface defect detection model of the circuit board;
an image acquisition module: the method comprises the steps of obtaining a surface image of a circuit board to be detected;
a pretreatment unit: cutting and normalizing the surface image;
a detection module: taking the preprocessed surface image as an input image of a detection model, detecting the input image by the detection model, adjusting the detected image into an output image of the original circuit board in size, and simultaneously outputting coordinate information, defect types and defect quantity of defect positions in the output image;
a defect marking module: and the manipulator is used for controlling the marking of the defect position.
Furthermore, the image acquisition module is provided with a matrix camera, an illumination unit, a cantilever support and an image acquisition card;
the matrix camera and the illumination unit are fixed on the cantilever support, and the image acquisition card is used for storing images acquired by the matrix camera.
Furthermore, the matrix camera is a CCD industrial camera, and the illumination unit consists of shadowless LED lamps.
Further, the detection module has a computer storage disk, a deep learning model, a GPU image processor, and a post-detection image marker display.
Furthermore, the detection system is also provided with a feeding conveying module, and the feeding conveying module is provided with a conveyor belt, a circuit board positioning device and a motion control unit;
the motion control unit is used for controlling the opening or closing of the conveyor belt, and the circuit board positioning device is used for positioning the circuit board after rotating and positioning;
the conveyor belt is used for conveying the circuit board.
Further, the defect marking module comprises a manipulator, a marking pen arranged on the manipulator and a movement control unit.
Compared with the prior art, the invention has the following advantages and effects:
the circuit board final inspection station detection method can improve the detection speed and the identification precision of the circuit board final inspection station, save labor cost for production enterprises, and simultaneously avoid the risk of occurrence of defects such as re-scratching and the like caused by a manual inspection link;
the invention uses a Faster R-CNN target detection technology based on Tensorflow improvement, different types of defects such as broken lines, sundries, scratches, dirt, poor repaired lines, hole loss, line defect, excessive oil supplement and the like which are suitable for detecting the surfaces of circuit boards of different varieties and different sizes can be obtained after training on a training data set, the detection speed can identify 14 images per second, the expansibility is strong, and simultaneously, the number of each defect type can be rapidly counted, so that the source of problems can be found as soon as possible, and the defects can be prevented from being generated again in time;
the scheme of accurately marking the defect position by using the manipulator can provide reference for subsequent circuit board defect repair and circuit board scrapping, and is favorable for improving the efficiency of repairing the surface defect of the circuit board and the efficiency of repairing the scrapped circuit board.
Drawings
FIG. 1 is a flowchart of a deep learning-based method for detecting defects on a surface of a circuit board according to an embodiment of the present invention;
FIG. 2 is a block diagram of a circuit board surface defect detection system based on deep learning according to an embodiment of the present invention;
fig. 3 is a schematic structural diagram of the circuit board surface defect detection based on deep learning according to the embodiment of the present invention.
Wherein, in the embodiment of the invention: 1. a feeding and conveying module; 2. an image acquisition module; 3. a detection module; 4. a defect marking module; 41. a manipulator; 42. a marking pen.
Detailed Description
The present invention will be described in further detail with reference to examples and drawings, but the present invention is not limited thereto.
As shown in fig. 1 to 3, a method for detecting surface defects of a circuit board based on deep learning includes the following steps: obtaining a detection model of the surface defects of the circuit board; the method specifically comprises the following steps: firstly, acquiring an original image of a surface defect of a standard circuit board; the standard circuit board is a circuit board which has defects in the prior art.
Preprocessing an original image; the method comprises the steps of realizing random rotation, cutting and normalization preprocessing of images on a computer based on a Python programming algorithm.
Labeling the preprocessed image by using labelImg and establishing a circuit board defect image data set; the circuit board defect image dataset is used for training deep learning. The circuit board defect image data set comprises defect images of broken lines, sundries, scratches, dirt, poor repaired lines, hole loss, line defect, overproof oil repair and the like on the surface of the circuit board. In this embodiment, there are 31348 total types of defect images of circuit boards, such as broken lines, impurities, scratches, dirt, defective repaired lines, hole defects, line defects, and excessive oil repair, on the surfaces of circuit boards of different types and sizes.
Building an improved Faster R-CNN target detection model based on Tensorflow;
a target detection model is trained on the circuit board defect image dataset data.
An improved Faster R-CNN target detection model based on Tenflow is built on a computer provided with an Ubuntu16.04 system, a backbone network uses an improved Resnet101 network, the model is trained on a labeled circuit board defect image data set, and model parameters suitable for detecting the surface defects of various circuit boards are obtained.
Acquiring a surface image of a circuit board to be detected;
preprocessing the surface image; this step includes image cropping and normalization based on Python programming algorithms on a computer.
And taking the preprocessed surface image as an input image of a detection model, detecting the input image by the detection model, adjusting the detected image into an output image of the original circuit board in size, and simultaneously outputting coordinate information, defect types and defect quantity of the defect position in the output image.
The image acquisition module 2 is used for completing the acquisition of the image of the circuit board to be detected, and the image is stored in a computer disk and is used as an input image of the image preprocessing module; then, image cropping is achieved on the computer based on a Python programming algorithm, and the cropped image is used as a detection input image for depth learning. And taking the image to be detected output in the last step as a trained input image of a Tensorflow improved based Faster R-CNN target detection model, splicing the detected output images into an image of the size of the original circuit board, and outputting the coordinate information of the defect position, the defect type and the defect number.
And controlling the mechanical arm to mark the defect position. And converting the coordinate information of the defect position output in the last step into positioning mark position information (x1, y1, z1, x2, y2 and z2) for controlling the manipulator, and separating the circuit board with the defect mark to the area to be repaired.
Another embodiment of the present invention further discloses a detection system using the deep learning-based circuit board surface defect detection method, including:
a model construction unit: the detection model is used for constructing the surface defect detection model of the circuit board;
the image acquisition module 2: the method comprises the steps of obtaining a surface image of a circuit board to be detected;
a pretreatment unit: preprocessing the surface image;
the detection module 3: taking the preprocessed surface image as an input image of a detection model, detecting the input image by the detection model, adjusting the detected image into an output image of the original circuit board in size, and simultaneously outputting coordinate information, defect types and defect quantity of defect positions in the output image;
defect marking module 4: and the manipulator is used for controlling the marking of the defect position.
The image acquisition module 2 is provided with a matrix camera, an illumination unit, a cantilever support and an image acquisition card. The matrix camera and the illumination unit are fixed on the cantilever support, and the image acquisition card is used for storing images acquired by the matrix camera. The matrix camera is a CCD industrial camera, and the illumination unit consists of shadowless LED lamps. The CCD industrial camera is composed of a CCD camera with model number FL2G-50S 5M.
The detection module 3 is provided with a computer storage disk, a deep learning model, a GPU image processor and a detected image mark display. The computer storage disk is a solid state disk with the size of 500G, the GPU image processor is an Nvidia GeForce GTX1060Ti 3G original works public display card, and the display model is HD Graphics 3000.
The detection system is also provided with a feeding conveying module 1, and the feeding conveying module 1 is provided with a conveyor belt, a circuit board positioning device and a motion control unit; the motion control unit is used for controlling the opening or closing of the conveyor belt, and the circuit board positioning device is used for positioning the circuit board after rotating and positioning; the conveyor belt is used for conveying the circuit board.
The defect marking module 4 includes a robot 41, a marker 42 provided on the robot 41, and a movement control unit. The manipulator has a horizontal X-axis and a horizontal Y-axis and a vertical Z-axis (wherein the positive direction of the X-axis is the advancing direction of the conveying belt) three-degree-of-freedom moving direction, the origin of coordinates is the position of the central point of the circuit board to be detected, and the manipulator positioning control module consists of a driving motor and a detection device.
The method for automatically detecting the surface defects of the circuit board based on deep learning comprises the following specific implementation steps:
(1) aiming at the circuit board with defects in the prior period:
obtaining a defect image of the surface of a circuit board and preprocessing the defect image
Firstly, an image acquisition module 2 is used for completing the acquisition of a circuit board defect image, and the circuit board defect image is stored in a computer disk and is used as an input image of an image preprocessing module;
then, image cropping and random rotation are implemented on the computer based on Python programming algorithm.
Labeling the preprocessed image by using labelImg and establishing a circuit board defect image data set for deep learning training;
the data set comprises 31348 defect images of various types of circuit boards with different types and sizes, such as broken lines, sundries, scratches, dirt, poor line repairing, hole damage, line defect, excessive oil repairing and the like on the surfaces of the circuit boards.
Training deep learning model
An improved Faster R-CNN target detection model based on Tenflow is built on a computer provided with an Ubuntu16.04 system, a backbone network uses an improved Resnet101 network, the model is trained on a labeled circuit board defect image data set, and model parameters suitable for detecting the surface defects of various circuit boards are obtained.
(2) Aiming at the circuit board to be tested:
acquiring and preprocessing a surface image of a circuit board to be detected
The method comprises the steps that an image acquisition module is used for acquiring an image of a circuit board to be detected, and the image is stored in a computer disk and used as an input image of an image preprocessing module; then, image cropping is achieved on the computer based on a Python programming algorithm, and the cropped image is used as a detection input image for depth learning.
Detecting surface defects of circuit board based on deep learning technology
And taking the image to be detected output in the last step as a trained input image of a Faster R-CNN target detection model improved based on Tensorflow, splicing the detected output image into an image with the size of the original circuit board as shown in FIG. 3, and outputting the coordinate information of the defect position, the defect type and the defect number.
Accurate marking of defective locations on a circuit board surface
And converting the coordinate information of the defect position output in the last step into positioning mark position information (x1, y1, z1, x2, y2 and z2) for controlling the manipulator, and separating the circuit board with the defect mark to the area to be repaired.
The above embodiments are preferred embodiments of the present invention, but the present invention is not limited to the above embodiments, and any other changes, modifications, substitutions, combinations, and simplifications which do not depart from the spirit and principle of the present invention should be construed as equivalents thereof, and all such changes, modifications, substitutions, combinations, and simplifications are intended to be included in the scope of the present invention.
Claims (10)
1. A circuit board surface defect detection method based on deep learning is characterized by comprising the following steps:
obtaining a detection model of the surface defects of the circuit board;
acquiring a surface image of a circuit board to be detected;
preprocessing the surface image;
taking the preprocessed surface image as an input image of a detection model, detecting the input image by the detection model, adjusting the detected image into an output image of the original circuit board in size, and simultaneously outputting coordinate information, defect types and defect quantity of defect positions in the output image;
and controlling the mechanical arm to mark the defect position.
2. The circuit board surface defect detection method based on deep learning of claim 1, wherein obtaining the detection model of the circuit board surface defects comprises the following steps:
acquiring an original image of a surface defect of a standard circuit board;
carrying out random rotation, cutting and normalization pretreatment on an original image;
labeling the preprocessed image by using labelImg and establishing a circuit board defect image data set;
building an improved Faster R-CNN target detection model based on Tensorflow;
a target detection model is trained on the circuit board defect image dataset data.
3. The method for detecting the surface defects of the circuit board based on the deep learning as claimed in claim 2, wherein the circuit board defect image dataset comprises defect images of broken lines, sundries, scratches, dirt, poor repaired lines, hole loss, line defects, overproof oil repair and the like on the surface of the circuit board.
4. The deep learning-based circuit board surface defect detection method of claim 1, wherein the pre-processing of the surface image comprises clipping and normalizing the original image based on a Python programming algorithm.
5. An inspection system using the deep learning based circuit board surface defect inspection method according to any one of claims 1 to 4, comprising:
a model construction unit: the detection model is used for constructing the surface defect detection model of the circuit board;
an image acquisition module: the method comprises the steps of obtaining a surface image of a circuit board to be detected;
a pretreatment unit: cutting and normalizing the surface image;
a detection module: taking the preprocessed surface image as an input image of a detection model, detecting the input image by the detection model, adjusting the detected image into an output image of the original circuit board in size, and simultaneously outputting coordinate information, defect types and defect quantity of defect positions in the output image;
a defect marking module: and the manipulator is used for controlling the marking of the defect position.
6. The detection system of the deep learning based circuit board surface defect detection method according to claim 5, wherein the image acquisition module has a matrix camera, an illumination unit, a cantilever support and an image acquisition card;
the matrix camera and the illumination unit are fixed on the cantilever support, and the image acquisition card is used for storing images acquired by the matrix camera.
7. The detection system of the deep learning based circuit board surface defect detection method according to claim 6, wherein the matrix camera is a CCD industrial camera, and the illumination unit is composed of shadowless LED lamps.
8. The system for detecting defects on surface of circuit board based on deep learning of claim 5, wherein the detection module comprises a computer storage disk, a deep learning model, a GPU image processor and a display of post-detection image markers.
9. The detection system of the deep learning based circuit board surface defect detection method according to claim 5, wherein the detection system further comprises a feeding and conveying module, the feeding and conveying module comprises a conveyor belt, a circuit board positioning device and a motion control unit;
the motion control unit is used for controlling the opening or closing of the conveyor belt, and the circuit board positioning device is used for positioning the circuit board after rotating and positioning;
the conveyor belt is used for conveying the circuit board.
10. The system for detecting defects on the surface of a circuit board based on deep learning of claim 5, wherein the defect marking module comprises a manipulator, a marker pen arranged on the manipulator and a movement control unit.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010401659.3A CN111768363A (en) | 2020-05-13 | 2020-05-13 | Deep learning-based circuit board surface defect detection method and detection system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010401659.3A CN111768363A (en) | 2020-05-13 | 2020-05-13 | Deep learning-based circuit board surface defect detection method and detection system |
Publications (1)
Publication Number | Publication Date |
---|---|
CN111768363A true CN111768363A (en) | 2020-10-13 |
Family
ID=72719124
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010401659.3A Pending CN111768363A (en) | 2020-05-13 | 2020-05-13 | Deep learning-based circuit board surface defect detection method and detection system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111768363A (en) |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112365491A (en) * | 2020-11-27 | 2021-02-12 | 上海市计算技术研究所 | Method for detecting welding seam of container, electronic equipment and storage medium |
CN112581462A (en) * | 2020-12-25 | 2021-03-30 | 北京邮电大学 | Method and device for detecting appearance defects of industrial products and storage medium |
CN112816484A (en) * | 2020-12-30 | 2021-05-18 | 重庆凯歌电子股份有限公司 | Detection system for PCB |
CN113222913A (en) * | 2021-04-28 | 2021-08-06 | 南京南瑞继保电气有限公司 | Circuit board defect detection positioning method and device and storage medium |
CN113533348A (en) * | 2021-07-16 | 2021-10-22 | 陕西科技大学 | Circuit board defect detection system based on deep learning and detection method thereof |
CN113822882A (en) * | 2021-11-22 | 2021-12-21 | 武汉飞恩微电子有限公司 | Circuit board surface defect detection method and device based on deep learning |
CN114037700A (en) * | 2021-12-31 | 2022-02-11 | 成都数联云算科技有限公司 | Panel defect detection method and device, electronic equipment and storage medium |
CN114494142A (en) * | 2021-12-28 | 2022-05-13 | 深圳科瑞技术股份有限公司 | Mobile terminal middle frame defect detection method and device based on deep learning |
CN115239712A (en) * | 2022-09-21 | 2022-10-25 | 季华实验室 | Circuit board surface defect detection method and device, electronic equipment and storage medium |
CN115684172A (en) * | 2022-10-09 | 2023-02-03 | 迁安市福运机动车检测有限公司 | Automobile appearance detection system and using method thereof |
CN116674300A (en) * | 2023-05-24 | 2023-09-01 | 常州润来科技有限公司 | Automatic ink-jet marking system and method based on copper pipe flaw detection result |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20010073376A (en) * | 2000-01-14 | 2001-08-01 | 윤종용 | Defective sign marking device for printed circuit board |
KR100673962B1 (en) * | 2005-07-22 | 2007-01-24 | 삼성테크윈 주식회사 | Apparatus for automatic optical inspection of flexible printed circuit board |
CN108921056A (en) * | 2018-06-18 | 2018-11-30 | 上海大学 | Pedestrian detection method based on neural network towards automobile assistant driving |
CN109636772A (en) * | 2018-10-25 | 2019-04-16 | 同济大学 | The defect inspection method on the irregular shape intermetallic composite coating surface based on deep learning |
CN109948563A (en) * | 2019-03-22 | 2019-06-28 | 华南农业大学 | A kind of withered tree detection localization method of the pine nematode based on deep learning |
CN110726735A (en) * | 2019-09-03 | 2020-01-24 | 北京精思博智科技有限公司 | Full-automatic circuit board defect detection system and method based on deep learning |
-
2020
- 2020-05-13 CN CN202010401659.3A patent/CN111768363A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20010073376A (en) * | 2000-01-14 | 2001-08-01 | 윤종용 | Defective sign marking device for printed circuit board |
KR100673962B1 (en) * | 2005-07-22 | 2007-01-24 | 삼성테크윈 주식회사 | Apparatus for automatic optical inspection of flexible printed circuit board |
CN108921056A (en) * | 2018-06-18 | 2018-11-30 | 上海大学 | Pedestrian detection method based on neural network towards automobile assistant driving |
CN109636772A (en) * | 2018-10-25 | 2019-04-16 | 同济大学 | The defect inspection method on the irregular shape intermetallic composite coating surface based on deep learning |
CN109948563A (en) * | 2019-03-22 | 2019-06-28 | 华南农业大学 | A kind of withered tree detection localization method of the pine nematode based on deep learning |
CN110726735A (en) * | 2019-09-03 | 2020-01-24 | 北京精思博智科技有限公司 | Full-automatic circuit board defect detection system and method based on deep learning |
Non-Patent Citations (1)
Title |
---|
姚建华 等: "全卷积神经网络下的多光谱遥感影像分割", 《中国图像图形学报》, vol. 25, no. 1, 31 January 2020 (2020-01-31), pages 180 - 192 * |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112365491A (en) * | 2020-11-27 | 2021-02-12 | 上海市计算技术研究所 | Method for detecting welding seam of container, electronic equipment and storage medium |
CN112581462A (en) * | 2020-12-25 | 2021-03-30 | 北京邮电大学 | Method and device for detecting appearance defects of industrial products and storage medium |
CN112816484A (en) * | 2020-12-30 | 2021-05-18 | 重庆凯歌电子股份有限公司 | Detection system for PCB |
CN113222913A (en) * | 2021-04-28 | 2021-08-06 | 南京南瑞继保电气有限公司 | Circuit board defect detection positioning method and device and storage medium |
CN113222913B (en) * | 2021-04-28 | 2024-04-12 | 南京南瑞继保电气有限公司 | Circuit board defect detection positioning method, device and storage medium |
CN113533348A (en) * | 2021-07-16 | 2021-10-22 | 陕西科技大学 | Circuit board defect detection system based on deep learning and detection method thereof |
CN113822882A (en) * | 2021-11-22 | 2021-12-21 | 武汉飞恩微电子有限公司 | Circuit board surface defect detection method and device based on deep learning |
CN114494142A (en) * | 2021-12-28 | 2022-05-13 | 深圳科瑞技术股份有限公司 | Mobile terminal middle frame defect detection method and device based on deep learning |
CN114037700A (en) * | 2021-12-31 | 2022-02-11 | 成都数联云算科技有限公司 | Panel defect detection method and device, electronic equipment and storage medium |
CN115239712A (en) * | 2022-09-21 | 2022-10-25 | 季华实验室 | Circuit board surface defect detection method and device, electronic equipment and storage medium |
CN115684172A (en) * | 2022-10-09 | 2023-02-03 | 迁安市福运机动车检测有限公司 | Automobile appearance detection system and using method thereof |
CN116674300A (en) * | 2023-05-24 | 2023-09-01 | 常州润来科技有限公司 | Automatic ink-jet marking system and method based on copper pipe flaw detection result |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111768363A (en) | Deep learning-based circuit board surface defect detection method and detection system | |
CN111007073B (en) | Method and system for online detection of part defects in additive manufacturing process | |
KR20220042916A (en) | Vision inspection system by using remote learning of product defects image | |
CN111982921B (en) | Method and device for detecting hole defects, conveying platform and storage medium | |
CN107966454A (en) | A kind of end plug defect detecting device and detection method based on FPGA | |
CN113418933B (en) | Flying shooting visual imaging detection system and method for detecting large-size object | |
CN113221889A (en) | Anti-interference recognition method and device for chip characters | |
CN108844966A (en) | A kind of screen detection method and detection system | |
CN113237889A (en) | Multi-scale ceramic detection method and system | |
CN111551559A (en) | LCD (liquid Crystal display) liquid crystal screen defect detection method based on multi-view vision system | |
CN112718552A (en) | Automatic quality defect inspection device for LED circuit board and working method thereof | |
CN112102254A (en) | Wood surface defect detection method and system based on machine vision | |
CN111929239A (en) | AOI detection device and detection method for PCB part defects | |
CN115493843A (en) | Quality monitoring method and equipment based on bearing retainer | |
CN110940672A (en) | Automatic generation method and device for intelligent optical detection sample characteristic and flaw AI model | |
JP6823156B2 (en) | Backup pin recognition method and component mounting device | |
CN112415013A (en) | Copper foil defect detection system | |
CN112763496A (en) | Mobile phone battery surface defect detection device and detection method thereof | |
US11327028B2 (en) | Ceramic ball automatic sorting system and method | |
CN106841211A (en) | Platform and method of a kind of utilization machine vision to cell piece surface defects detection | |
JP2015137921A (en) | Appearance inspection device, appearance inspection method, and program | |
CN113495073A (en) | Auto-focus function for vision inspection system | |
CN111272776A (en) | Flexible circuit board defect detection method based on AI technology | |
CN216208660U (en) | Plastic bottle body defect detection device | |
CN108705689B (en) | Automatic tool setting process method for repairing surface microdefects of large-caliber precise optical crystal |
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 |