CN107886131A - One kind is based on convolutional neural networks detection circuit board element polarity method and apparatus - Google Patents
One kind is based on convolutional neural networks detection circuit board element polarity method and apparatus Download PDFInfo
- Publication number
- CN107886131A CN107886131A CN201711189432.1A CN201711189432A CN107886131A CN 107886131 A CN107886131 A CN 107886131A CN 201711189432 A CN201711189432 A CN 201711189432A CN 107886131 A CN107886131 A CN 107886131A
- Authority
- CN
- China
- Prior art keywords
- circuit board
- polarity
- camera
- image
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 27
- 238000013527 convolutional neural network Methods 0.000 title claims abstract description 26
- 238000001514 detection method Methods 0.000 title claims abstract description 24
- 238000001914 filtration Methods 0.000 claims abstract description 16
- 238000005457 optimization Methods 0.000 claims abstract description 9
- 230000008569 process Effects 0.000 claims abstract description 8
- 238000005286 illumination Methods 0.000 claims description 4
- 230000008901 benefit Effects 0.000 abstract description 3
- 238000012549 training Methods 0.000 description 8
- 238000005516 engineering process Methods 0.000 description 5
- 239000000284 extract Substances 0.000 description 5
- 238000005070 sampling Methods 0.000 description 4
- 230000007547 defect Effects 0.000 description 3
- 238000007689 inspection Methods 0.000 description 3
- 238000013507 mapping Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 230000004913 activation Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 238000005520 cutting process Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 231100000518 lethal Toxicity 0.000 description 1
- 230000001665 lethal effect Effects 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computational Linguistics (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The present invention discloses a kind of based on convolutional neural networks detection circuit board element polarity method, adjust camera setting height(from bottom) and acquisition parameters and adjustment lighting source, choose certain amount circuit board and establish target element device image database of the polarity correctly with mistake, gained polarity is correct and the target element device image of the circuit board of incorrect polarity is trained optimization as the input picture of convolutional neural networks, obtain optimizing polar character grader, camera is eliminated the target element device image to be measured of noise to circuit board to be detected shooting and gaussian filtering process, polar character grader judges to obtain final detection result to filtered target element device image, target element device image after the judgement learning object as polar character grader again.Invention additionally discloses use the device based on convolutional neural networks detection circuit board element polarity method.Compared with the prior art, the present invention has the advantages that quick, accurate, reliable, can accelerate circuit board element detection speed.
Description
Technical field
It is more particularly to a kind of based on convolutional neural networks detection circuit board the present invention relates to mechanical vision inspection technology field
Component polarity method and apparatus.
Background technology
In recent years, as electronic technology continues to develop, printed circuit board (PCB) also obtains as the important component of electronic technology
To developing rapidly, but whether the polar orientation for having directive electronic component in circuit board is correctly decision-making circuit plate quality
One of factor.
During board production, it is understood that there may be various undesirable elements influence, and the part component on circuit board can
The true phenomenon of poor direction can occur, and this defect turns into the lethal factor of circuit board, influences the final mass of circuit board.
Therefore, it is to ensure quality of printed circuits, it is essential, and the wherein polarity of component that defects detection is carried out to circuit board
It is exactly one of defect.In existing production process, many producers can employ large quantities of people to enter pedestrian to the polarity of circuit board element
Work detects, but not only efficiency is low but also because worker lacks experience for artificial detection, is easily leaked in circuit board checking process
Phenomena such as inspection, false retrieval, therefore the final quality that circuit board can not also be completely secured.
In order to solve problem present in above-mentioned hand inspection circuit boards, Su Mingming master's thesis《Circuit
The detection and identification of board component》By the edge feature of edge extracting circuit board element, then extracted using hough transform
Linear feature is compared to judge the polarity of component with standard information again, but this method processing speed is slow, and error may
Property it is high.Therefore, using the polarity for going to detect circuit board element with reliable algorithm to improving detection efficiency with extremely heavy
Want meaning.
The content of the invention
The main object of the present invention is to propose that one kind is quick, precisely, reliably detects circuit board based on convolutional neural networks
Component polarity method, the present invention also propose a kind of use based on convolutional neural networks detection circuit board element polarity method
Device, it is intended to accelerate the speed of circuit board element detection.
To achieve the above object, it is proposed by the present invention a kind of based on convolutional neural networks detection circuit board element polarity side
Method, comprise the following steps：
S1：The setting height(from bottom) and acquisition parameters of camera are adjusted, while adjusts the intensity of illumination of lighting source；
S2：Choose the correct circuit board of certain amount polarity and establish the correct target element device image database of polarity；
S3：Choose the circuit board of certain amount incorrect polarity and establish the target element device image database of incorrect polarity；
S4：By the polarity obtained by the step S2 and the step S3 is correct and the target of the circuit board of incorrect polarity
Component image is trained optimization as the input picture of convolutional neural networks, the polar character grader optimized；
S5:The camera, which shoots to circuit board to be detected and carries out gaussian filtering process, obtains target component figure to be measured
Picture, the polar character grader judge to obtain final detection result to target element device image after filtering；
S6：The target element device image after the judgement learning object as the polar character grader again, enters one
Step optimization polar character grader.
Preferably, the step S2 further comprises the steps：
S21：The a number of correct circuit board of component polarity is transmitted by conveyer belt and by the camera shooting figure
Picture；
S22：The image that the camera is shot passes through gaussian filtering, eliminates picture noise；
S23：Selection puts most regular circuit board image as template, by the target element on each correct sampler plate
Device cuts out and preserved after carrying out corners Matching with target component in template.
Preferably, the step S3 further comprises the steps：
S31：The circuit board of a number of component incorrect polarity is transmitted by conveyer belt and by the camera shooting figure
Picture；
S32：By the image shot by camera after gaussian filtering, picture noise is eliminated；
S33：Selection puts most regular circuit board image as template, by the target element on each error sample circuit board
Device cuts out and preserved after carrying out corners Matching with target component in template.
Preferably, the circuit board quantity of the correct circuit board of polarity and incorrect polarity is 100.
The present invention also proposes the dress based on convolutional neural networks detection circuit board element polarity method described in a kind of use
Put, including the conveyer belt that circuit board is conveyed from one end to the other side, the conveyer belt top surface have been sequentially placed multiple circuit boards,
Camera is provided with above the conveyer belt also to electrically connect with display with calculating mechatronics, the computer.
Preferably, emitted beam between the camera and circuit board provided with lighting source and be irradiated in circuit board surface.
Preferably, the camera is CCD industrial cameras.
Technical solution of the present invention has advantages below compared with the prior art：
Technical solution of the present invention uses convolutional neural networks scheme, by a collection of polarity is correct and the member of incorrect polarity
Device carries out deep learning, trains the feature classifiers of optimization, and wherein training process need to only be carried out once, then passed through again
Feature classifiers are detected to the polarity of circuit board element to be detected to obtain result.
Technical solution of the present invention is very fast to the sweep speed of the component of circuit board under test, and stability is preferable, can not only
Improve the processing speed of processing target component so that Detection results are more preferable, and algorithm is also more stable.
Brief description of the drawings
In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing
There is the required accompanying drawing used in technology description to be briefly described, it should be apparent that, drawings in the following description are only this
Some embodiments of invention, for those of ordinary skill in the art, on the premise of not paying creative work, can be with
Structure according to these accompanying drawings obtains other accompanying drawings.
Fig. 1 is the target element units test flow chart of the present invention；
Fig. 2 is the schematic diagram of the convolutional neural networks training of the present invention；
Fig. 3 is the apparatus structure schematic diagram of the present invention.
Drawing reference numeral explanation：
| Label | Title | Label | Title |
| 1 | Conveyer belt | 4 | Camera |
| 2 | Circuit board | 5 | Display |
| 3 | Lighting source | 6 | Computer |
The realization, functional characteristics and advantage of the object of the invention will be described further referring to the drawings in conjunction with the embodiments.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Site preparation describes, it is clear that described embodiment is only the part of the embodiment of the present invention, rather than whole embodiments.Base
Embodiment in the present invention, those of ordinary skill in the art obtained under the premise of creative work is not made it is all its
His embodiment, belongs to the scope of protection of the invention.
If it is to be appreciated that related in the embodiment of the present invention directionality instruction (such as up, down, left, right, before and after ...),
Then directionality instruction be only used for explaining relative position relation under a certain particular pose (as shown in drawings) between each part,
Motion conditions etc., if the particular pose changes, directionality instruction also correspondingly changes therewith.
If in addition, relating to the description of " first ", " second " etc. in the embodiment of the present invention, " first ", " second " etc. are somebody's turn to do
Description be only used for describing purpose, and it is not intended that instruction or implying its relative importance or implicit indicating indicated skill
The quantity of art feature.Thus, " first " is defined, the feature of " second " can be expressed or implicitly includes at least one spy
Sign.In addition, the technical scheme between each embodiment can be combined with each other, but must be with those of ordinary skill in the art's energy
Based on enough realizations, the knot of this technical scheme is will be understood that when the combination appearance of technical scheme is conflicting or can not realize
Conjunction is not present, also not within the protection domain of application claims.
The present invention proposes a kind of based on convolutional neural networks detection circuit board element polarity method.
Fig. 1 to Fig. 3 please be participate in, technical solution of the present invention detects circuit board element polarity side based on convolutional neural networks
Method comprises the following steps：
S1：The setting height(from bottom) and acquisition parameters of camera 4 are adjusted, while adjusts the intensity of illumination of lighting source 3；
S2：Choose the correct circuit board 2 of certain amount polarity and establish the correct target element device image database of polarity；
S3：Choose the circuit board 2 of certain amount incorrect polarity and establish the target element device image database of incorrect polarity；
S4：By obtained by step S2 and step S3 polarity correctly and incorrect polarity circuit board target element device image work
Optimization is trained for the input picture of convolutional neural networks, the polar character grader optimized；
S5:Camera 4, which shoots to circuit board 2 to be detected and carries out gaussian filtering process, obtains target element device image to be measured,
Polar character grader judges to obtain final detection result to target element device image after filtering；
S6：Learning object of the target element device image again as polar character grader after judgement further optimizes polarity
Feature classifiers.
In technical solution of the present invention, above-mentioned steps S2 further comprises the steps：
S21：The correct circuit board 2 of a number of component polarity is transmitted and by the shooting figure of camera 4 by conveyer belt 1
Picture；
S22：The image that camera 4 is shot passes through gaussian filtering, eliminates picture noise；
S23：Selection puts the most regular image of circuit board 2 as template, by the target on each correct sampler plate 2
Component cuts out and preserved after carrying out corners Matching with target component in template.
And above-mentioned steps S3 further comprises the steps：
S31：The circuit board 2 of a number of component incorrect polarity is transmitted and by the shooting figure of camera 4 by conveyer belt 1
Picture；
S32：By the shooting image of camera 4 after gaussian filtering, picture noise is eliminated；
S33：Selection puts the most regular image of circuit board 2 as template, by the target on each error sample circuit board 2
Component cuts out and preserved after carrying out corners Matching with target component in template.
Fig. 3 is referred to, the present invention also proposes a kind of using based on convolutional neural networks detection circuit board element polarity side
The device of method, including the conveyer belt 1 that circuit board 2 is conveyed from one end to the other side, the top surface of conveyer belt 1 have been sequentially placed multiple electricity
Road plate 2, the top of conveyer belt 1 electrically connect provided with camera 4 with computer 6, and computer 6 also electrically connects with display 5.Preferably, phase
Emitted beam between machine 4 and circuit board 2 provided with lighting source 3 and be irradiated in the surface of circuit board 2, camera 4 is CCD industrial cameras.
Refer to Fig. 1 to Fig. 3, in technical solution of the present invention, circuit board element is detected using based on convolutional neural networks
The operation principle of the device of polarity method is：
The setting height(from bottom) and acquisition parameters of camera 4 are adjusted first, and the illumination at the same time adjusting lighting source 3 is strong
Degree.Then the correct circuit board 2 of certain amount polarity, such as the selection correct circuit board 2 of 100 polarity are chosen, passes through transmission
Band 1 is transmitted to the lower section of camera 4 so that camera 4 is shot to circuit board 2, and camera 4 will shoot obtained image through too high
After this filtering, selection wherein puts the image of circuit board 2 the most regular as template, and then target component is cut,
Other circuit boards 2 are then by with intactly cutting the target component on each circuit board 2 after template progress corners Matching
Out, then it is stored in a file and saves.
In the same manner, the circuit board 2 of certain amount incorrect polarity, such as the circuit of 100 incorrect polarities of selection are then chosen
Plate 2, it is transmitted by conveyer belt 1 to the lower section of camera 4 so that camera 4 is shot to circuit board 2, and camera 4 obtains shooting
Image after gaussian filtering, selection wherein put the image of circuit board 2 the most regular as template, then to target element device
Part is cut, and other circuit boards 2 with template then by carrying out the target component on each circuit board 2 after corners Matching
Intactly cut out, be then stored in another file and save.
Then the target element device image of the correct circuit board that obtains above-mentioned steps and the circuit board of mistake as
The input picture of convolutional neural networks is trained optimization, so as to the polar character grader optimized.
In specific training optimization process, convolutional neural networks mainly extract stage and grader group by multilayer feature
Into the multilayer feature extraction stage usually contains 1~3 layer of convolutional layer and down-sampling layer, and grader has also generally comprised 1~2 layer
Full articulamentum.As shown in Fig. 2 the target element device image as training sample completes input sample image by input layer first
Step, and then convolutional layer.
As shown in Fig. 2 convolutional layer C is polar character extract layer, the input of each neuron and the local experiences of preceding layer
Open country is connected, and receptive field is the region of input picture corresponding to the response of some node of output type characteristic pattern.Local receptor field with
The training sample component image of input carries out convolution, extracts the local feature of circuit board element polarity.Then down-sampling layer S
It is polar character mapping layer, the polar character of input is compressed, so that polar character figure diminishes, and simplifies network calculations
Complexity, while polar character compression is carried out, main polar character is extracted, what Feature Mapping structure was checked and write off using influence function
Activation primitive of the sigmoid functions as convolutional neural networks so that polar character mapping has shift invariant.
In technical solution of the present invention, input treats that the size of training objective component image is m × n, and wherein k × k is convolution
The size of core, i.e. weight, and receptive field.It is (m-k) × (n-k) to obtain polarity output characteristic figure size by convolution, output
Polar character figure can pass through the sampling that size is w × h again, and can obtain size isPole
Property characteristic image.
First time convolution as shown in Figure 2, the convolution kernel of one 11 × 11 × 3 and 219 × 219 × 3 polarized
The input picture of component carries out convolution, and the polar character figure size of output is (219-11) × (219-11), and input size is
208 × 208 polar character figure is sampled by 4 × 4 size, with obtaining 53 × 53 polarity output characteristic.Then
Convolutional layer further extracts with down-sampling layer to polar character to be optimized, and training can obtain multiple polarity the most outstanding after terminating special
Levy grader.
After the component convolutional neural networks learning process for completing training sample, it is possible to target component to be detected
Examinations.On streamline, camera shooting circuit board image inputs computer and handled by gaussian filtering, but because
The inclination of putting of circuit board causes the position of target component to be measured not know, and the present invention utilizes corner correspondence by circuit
Plate is corrected to the putting position of standard, you can the corresponding relation of the feature pixel between two images is found, so that it is determined that two
The position relationship of width image, after image puts standard, the position of target component to be detected is plucked out using mouse, recycles position
Target component to be measured is taken out, extracts the classification that target component passes through feature classifiers, that is, according to convolutional Neural net
The learning experience of network is judged, exports " polarity is correct " or " the wrong mistake of polarity " of testing result.At the same time, above-mentioned detection
The image of the intermediate objective component of step again can be as component to be learned, so as to further allow feature classifiers more
Optimization.
The preferred embodiments of the present invention are the foregoing is only, are not intended to limit the scope of the invention, it is every at this
Under the design of invention, the equivalent structure transformation made using description of the invention and accompanying drawing content, or directly/it is used in indirectly
He is included in the scope of patent protection of the present invention related technical field.
Claims (7)
1. one kind is based on convolutional neural networks detection circuit board element polarity method, it is characterised in that comprises the following steps：
S1：The setting height(from bottom) and acquisition parameters of camera are adjusted, while adjusts the intensity of illumination of lighting source；
S2：Choose the correct circuit board of certain amount polarity and establish the correct target element device image database of polarity；
S3：Choose the circuit board of certain amount incorrect polarity and establish the target element device image database of incorrect polarity；
S4：By the polarity obtained by the step S2 and the step S3 is correct and the target element device of the circuit board of incorrect polarity
Part image is trained optimization as the input picture of convolutional neural networks, the polar character grader optimized；
S5:The camera, which shoots to circuit board to be detected and carries out gaussian filtering process, obtains target element device image to be measured, institute
State polar character grader target element device image after filtering is judged to obtain final detection result；
S6：The target element device image after the judgement learning object as the polar character grader again, it is further excellent
Change polar character grader.
2. the method as described in claim 1, it is characterised in that the step S2 further comprises the steps：
S21：The a number of correct circuit board of component polarity is transmitted by conveyer belt and by the image shot by camera；
S22：The image that the camera is shot passes through gaussian filtering, eliminates picture noise；
S23：Selection puts most regular circuit board image as template, by the target component on each correct sampler plate
Cut out and preserved after carrying out corners Matching with target component in template.
3. method as claimed in claim 2, it is characterised in that the step S3 further comprises the steps：
S31：The circuit board of a number of component incorrect polarity is transmitted by conveyer belt and by the image shot by camera；
S32：The image shot by camera is passed through into gaussian filtering, eliminates picture noise；
S33：Selection puts most regular circuit board image as template, by the target component on each error sample circuit board
Cut out and preserved after carrying out corners Matching with target component in template.
4. method as claimed in claim 3, it is characterised in that the circuit board quantity of the correct circuit board of polarity and incorrect polarity
It is 100.
A kind of 5. dress using based on convolutional neural networks detection circuit board element polarity method as described in Claims 1-4
Put, it is characterised in that including the conveyer belt for conveying circuit board from one end to the other side, the conveyer belt top surface has been sequentially placed
Multiple circuit boards, the conveyer belt top are provided with camera and also electrically connected with calculating mechatronics, the computer with display.
6. device as claimed in claim 5, it is characterised in that send light provided with lighting source between the camera and circuit board
Line is irradiated in circuit board surface.
7. device as claimed in claim 6, it is characterised in that the camera is CCD industrial cameras.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711189432.1A CN107886131A (en) | 2017-11-24 | 2017-11-24 | One kind is based on convolutional neural networks detection circuit board element polarity method and apparatus |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711189432.1A CN107886131A (en) | 2017-11-24 | 2017-11-24 | One kind is based on convolutional neural networks detection circuit board element polarity method and apparatus |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107886131A true CN107886131A (en) | 2018-04-06 |
Family
ID=61775100
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711189432.1A Pending CN107886131A (en) | 2017-11-24 | 2017-11-24 | One kind is based on convolutional neural networks detection circuit board element polarity method and apparatus |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107886131A (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109635856A (en) * | 2018-11-29 | 2019-04-16 | 上海集成电路研发中心有限公司 | A kind of producing line defect image intelligent classification system and classification method |
| CN110517260A (en) * | 2019-08-30 | 2019-11-29 | 北京地平线机器人技术研发有限公司 | The detection method and device of circuit board, storage medium, electronic equipment |
| CN110967851A (en) * | 2019-12-26 | 2020-04-07 | 成都数之联科技有限公司 | Circuit extraction method and system for array image of liquid crystal panel |
| CN111223077A (en) * | 2019-12-30 | 2020-06-02 | 浙江力创自动化科技有限公司 | Test method and test system of circuit board |
| CN112381751A (en) * | 2020-07-07 | 2021-02-19 | 昆山新精度金属科技有限公司 | Online intelligent detection system and method based on image processing algorithm |
| CN112967224A (en) * | 2021-01-29 | 2021-06-15 | 绍兴隆芙力智能科技发展有限公司 | Electronic circuit board detection system, method and medium based on artificial intelligence |
| CN113837209A (en) * | 2020-06-23 | 2021-12-24 | 乐达创意科技股份有限公司 | Method and system for improved machine learning using data for training |
| CN114202515A (en) * | 2021-11-29 | 2022-03-18 | 广州海谷电子科技有限公司 | Method for detecting defect of printed carbon line of humidity sensor |
| CN114428082A (en) * | 2020-10-29 | 2022-05-03 | 技嘉科技股份有限公司 | Electronic component image capturing method and capacitor polarity determination method using same |
| CN114638792A (en) * | 2022-03-03 | 2022-06-17 | 浙江达峰科技有限公司 | Method for detecting polarity defect of electrolytic capacitor of plug-in circuit board |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102539441A (en) * | 2011-12-29 | 2012-07-04 | 中国科学院长春光学精密机械与物理研究所 | Inspection method for inspecting welding correctness of the device of device on circuit board in batch |
| CN105426917A (en) * | 2015-11-23 | 2016-03-23 | 广州视源电子科技股份有限公司 | Component classification method and apparatus |
| CN105469400A (en) * | 2015-11-23 | 2016-04-06 | 广州视源电子科技股份有限公司 | Rapid identification and marking method of electronic component polarity direction and system thereof |
| CN105513046A (en) * | 2015-11-23 | 2016-04-20 | 广州视源电子科技股份有限公司 | Method and system for identifying electronic component poles, method and system for marking electronic component poles |
| CN106251333A (en) * | 2016-07-13 | 2016-12-21 | 广州视源电子科技股份有限公司 | Element anti-part detection method and system |
| CN106370671A (en) * | 2016-10-12 | 2017-02-01 | 浙江理工大学 | PCB (printed circuit board) component detection system and method based on machine vision |
| CN106778835A (en) * | 2016-11-29 | 2017-05-31 | 武汉大学 | The airport target by using remote sensing image recognition methods of fusion scene information and depth characteristic |
| CN106778740A (en) * | 2016-12-06 | 2017-05-31 | 北京航空航天大学 | A kind of TFDS non-faulting image detecting methods based on deep learning |
| CN107341805A (en) * | 2016-08-19 | 2017-11-10 | 北京市商汤科技开发有限公司 | Background segment and network model training, image processing method and device before image |
| CN107347125A (en) * | 2016-08-19 | 2017-11-14 | 北京市商汤科技开发有限公司 | Processing method, device and the terminal device of video image |
-
2017
- 2017-11-24 CN CN201711189432.1A patent/CN107886131A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102539441A (en) * | 2011-12-29 | 2012-07-04 | 中国科学院长春光学精密机械与物理研究所 | Inspection method for inspecting welding correctness of the device of device on circuit board in batch |
| CN105426917A (en) * | 2015-11-23 | 2016-03-23 | 广州视源电子科技股份有限公司 | Component classification method and apparatus |
| CN105469400A (en) * | 2015-11-23 | 2016-04-06 | 广州视源电子科技股份有限公司 | Rapid identification and marking method of electronic component polarity direction and system thereof |
| CN105513046A (en) * | 2015-11-23 | 2016-04-20 | 广州视源电子科技股份有限公司 | Method and system for identifying electronic component poles, method and system for marking electronic component poles |
| CN106251333A (en) * | 2016-07-13 | 2016-12-21 | 广州视源电子科技股份有限公司 | Element anti-part detection method and system |
| CN107341805A (en) * | 2016-08-19 | 2017-11-10 | 北京市商汤科技开发有限公司 | Background segment and network model training, image processing method and device before image |
| CN107347125A (en) * | 2016-08-19 | 2017-11-14 | 北京市商汤科技开发有限公司 | Processing method, device and the terminal device of video image |
| CN106370671A (en) * | 2016-10-12 | 2017-02-01 | 浙江理工大学 | PCB (printed circuit board) component detection system and method based on machine vision |
| CN106778835A (en) * | 2016-11-29 | 2017-05-31 | 武汉大学 | The airport target by using remote sensing image recognition methods of fusion scene information and depth characteristic |
| CN106778740A (en) * | 2016-12-06 | 2017-05-31 | 北京航空航天大学 | A kind of TFDS non-faulting image detecting methods based on deep learning |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109635856A (en) * | 2018-11-29 | 2019-04-16 | 上海集成电路研发中心有限公司 | A kind of producing line defect image intelligent classification system and classification method |
| CN110517260A (en) * | 2019-08-30 | 2019-11-29 | 北京地平线机器人技术研发有限公司 | The detection method and device of circuit board, storage medium, electronic equipment |
| CN110967851A (en) * | 2019-12-26 | 2020-04-07 | 成都数之联科技有限公司 | Circuit extraction method and system for array image of liquid crystal panel |
| CN110967851B (en) * | 2019-12-26 | 2022-06-21 | 成都数之联科技股份有限公司 | Line extraction method and system for array image of liquid crystal panel |
| CN111223077A (en) * | 2019-12-30 | 2020-06-02 | 浙江力创自动化科技有限公司 | Test method and test system of circuit board |
| CN113837209A (en) * | 2020-06-23 | 2021-12-24 | 乐达创意科技股份有限公司 | Method and system for improved machine learning using data for training |
| CN112381751A (en) * | 2020-07-07 | 2021-02-19 | 昆山新精度金属科技有限公司 | Online intelligent detection system and method based on image processing algorithm |
| CN114428082A (en) * | 2020-10-29 | 2022-05-03 | 技嘉科技股份有限公司 | Electronic component image capturing method and capacitor polarity determination method using same |
| CN112967224A (en) * | 2021-01-29 | 2021-06-15 | 绍兴隆芙力智能科技发展有限公司 | Electronic circuit board detection system, method and medium based on artificial intelligence |
| CN114202515A (en) * | 2021-11-29 | 2022-03-18 | 广州海谷电子科技有限公司 | Method for detecting defect of printed carbon line of humidity sensor |
| CN114638792A (en) * | 2022-03-03 | 2022-06-17 | 浙江达峰科技有限公司 | Method for detecting polarity defect of electrolytic capacitor of plug-in circuit board |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107886131A (en) | One kind is based on convolutional neural networks detection circuit board element polarity method and apparatus | |
| CN109239102B (en) | CNN-based flexible circuit board appearance defect detection method | |
| CN105469400B (en) | The quick method and system for identifying, marking of electronic component polar orientation | |
| CN107389701A (en) | A kind of PCB visual defects automatic checkout system and method based on image | |
| CN109959661A (en) | Automatic optical detection method, equipment and its deep learning system | |
| CN111899241B (en) | Quantitative on-line detection method and system for defects of PCB (printed Circuit Board) patches in front of furnace | |
| CN108109137A (en) | The Machine Vision Inspecting System and method of vehicle part | |
| CN109584227A (en) | A kind of quality of welding spot detection method and its realization system based on deep learning algorithm of target detection | |
| CN106651802B (en) | Machine vision scolding tin position finding and detection method | |
| CN106875381A (en) | A kind of phone housing defect inspection method based on deep learning | |
| CN105513046B (en) | The polar recognition methods of electronic component and system, mask method and system | |
| CN106875373A (en) | Mobile phone screen MURA defect inspection methods based on convolutional neural networks pruning algorithms | |
| CN109683360A (en) | Liquid crystal display panel defect inspection method and device | |
| CN106501272A (en) | Machine vision scolding tin position detecting system | |
| CN104867145B (en) | IC element welding point defect detection methods based on VIBE models | |
| CN104867144A (en) | IC element solder joint defect detection method based on Gaussian mixture model | |
| CN110415214A (en) | Appearance detecting method, device, electronic equipment and the storage medium of camera module | |
| CN110517265A (en) | A kind of detection method of surface defects of products, device and storage medium | |
| CN106355579A (en) | Defect detecting method of cigarette carton surface wrinkles | |
| CN113538375A (en) | PCB defect detection method based on YOLOv5 | |
| CN109946304A (en) | Surface defects of parts on-line detecting system and detection method based on characteristic matching | |
| CN113222938A (en) | Chip defect detection method and system and computer readable storage medium | |
| CN111929239A (en) | AOI detection device and detection method for PCB part defects | |
| CN105354816B (en) | A kind of electronic units fix method and device | |
| CN105023018A (en) | Jet code detection method and system |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180406 |