CN109871895A - The defect inspection method and device of circuit board - Google Patents
The defect inspection method and device of circuit board Download PDFInfo
- Publication number
- CN109871895A CN109871895A CN201910133359.9A CN201910133359A CN109871895A CN 109871895 A CN109871895 A CN 109871895A CN 201910133359 A CN201910133359 A CN 201910133359A CN 109871895 A CN109871895 A CN 109871895A
- Authority
- CN
- China
- Prior art keywords
- defect
- image
- pixel
- disaggregated model
- circuit
- 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
Abstract
The application proposes the defect inspection method and device of a kind of circuit board, wherein, method includes: the shooting image by obtaining circuit-under-test plate, image will be shot and standard picture respective pixel carries out Pixel Information difference, obtain the difference value of respective pixel, the difference value of the Pixel Information and respective pixel that shoot each pixel of image is synthesized, obtain input picture, input picture is inputted into trained disaggregated model, according to the output of disaggregated model, determine the defect information of circuit-under-test plate, wherein, disaggregated model has learnt to obtain the characteristics of image of the corresponding input picture of circuit board of all kinds of defects.The characteristics of image for the corresponding input picture of circuit board for obtaining all kinds of defects is learnt due to disaggregated model, therefore, image characteristics extraction is carried out in Pixel Information by the way that the difference value to be increased to shooting image, the input data that disaggregated model can be enriched helps to improve the precision and efficiency of the detection of circuit-under-test board defect.
Description
Technical field
This application involves deep learning and technical field of image processing more particularly to a kind of defect inspection methods of circuit board
And device.
Background technique
At present circuit board fabrication dependent on automation industrial production line, due to circuit board electronic component integrated level not
Disconnected to increase, board production technique becomes increasingly complex, and inevitably will appear defective circuits during circuit board generates
Therefore plate carries out defects detection, circuit board damage caused by can be avoided because of defect to circuit board.
In the prior art, circuit board, which generates corporate boss, to carry out defects detection to circuit board using artificial detection method.But
That artificial detection needs staff with the naked eye to check, there are testing costs it is high, accuracy is lower, low efficiency the disadvantages of.
Summary of the invention
The application is intended to solve at least some of the technical problems in related technologies.
The application proposes the defect inspection method and device of a kind of circuit board, solves in the prior art, to circuit board
When defects detection relies on artificial detection, there is technical issues that accuracy rate is lower, testing cost is high and.
The application first aspect embodiment proposes a kind of defect inspection method of circuit board, comprising:
Obtain the shooting image of circuit-under-test plate；
The shooting image and standard picture respective pixel are subjected to Pixel Information difference, obtain the difference of respective pixel
Value；Wherein, the standard picture is shot to the reference circuit plate of non-existing defects；
The difference value of the Pixel Information of shooting each pixel of image and respective pixel is synthesized, input figure is obtained
Picture；
The input picture is inputted into trained disaggregated model, to determine institute according to the output of the disaggregated model
State the defect information of circuit-under-test plate；Wherein, the disaggregated model learnt to obtain all kinds of defects circuit board it is corresponding defeated
Enter the characteristics of image of image.
It is each in the shooting image and the standard picture as the first possible implementation of the embodiment of the present application
The Pixel Information of pixel includes the gray value of depth value and each color channel.
It is described to shoot image picture corresponding with standard picture as second of possible implementation of the embodiment of the present application
Element carries out Pixel Information difference, obtains the difference value of respective pixel, comprising:
Respective pixel in the shooting image and the standard picture is subjected to depth value difference, obtains the depth of respective pixel
Spend difference value；
Respective pixel in the shooting image and the standard picture is subjected to gray value difference, obtains respective pixel each
The gray scale difference score value of color channel.
As the third possible implementation of the embodiment of the present application, the pixel by shooting each pixel of image
The difference value of information and respective pixel is synthesized, and input picture is obtained, comprising:
Using the gray scale difference score value of the depth difference value, the depth value and each color channel and gray value as institute
State the Pixel Information of respective pixel in input picture.
As the 4th kind of possible implementation of the embodiment of the present application, the disaggregated model includes convolutional layer, pond layer
With full articulamentum；
Wherein, the convolutional layer carries out image characteristics extraction for the Pixel Information to each pixel in the input picture；
The pond layer, the feature for extracting to the convolutional layer carry out dimensionality reduction operation；
The full articulamentum, for being classified according to the characteristics of image after the pond layer dimensionality reduction.
As the 5th kind of possible implementation of the embodiment of the present application, the disaggregated model is by sample image and institute
State the pixel that standard picture respective pixel carries out respective pixel in the difference value and the sample image that Pixel Information difference obtains
Information synthesis, is trained using the sample image after synthesis；Wherein, the sample image has marked defect information,
The defect information is used to indicate the circuit board of respective sample image displaying with the presence or absence of defect and defect type；
When the defect information difference of the defect information of disaggregated model output and sample image mark minimizes,
The disaggregated model training terminates.
It is described to input the input picture by training as the 6th kind of possible implementation of the embodiment of the present application
Disaggregated model, with according to the output of the disaggregated model, after the defect information for determining the circuit-under-test plate, further includes:
If the circuit-under-test plate existing defects, the shooting image is shown in control interface；
It obtains and the defect information that artificial defect type marks is carried out to the shooting image；
According to the shooting image and the defect information, the first training sample is generated；
The disaggregated model is trained using first training sample.
It is described to input the input picture by training as the 7th kind of possible implementation of the embodiment of the present application
Disaggregated model, with according to the output of the disaggregated model, after the defect information for determining the circuit-under-test plate, further includes:
Never selected part circuit board in the circuit-under-test plate of existing defects shows the partial circuit chosen in control interface
The shooting image of plate；
To the partial circuit plate of selection, the defect information that corresponding shooting image manually marks is obtained, to generate for institute
State the second training sample that disaggregated model is trained.
It is described to input the input picture by training as the 8th kind of possible implementation of the embodiment of the present application
Disaggregated model, with according to the output of the disaggregated model, after the defect information for determining the circuit-under-test plate, further includes:
If the circuit-under-test plate existing defects, control production line, the circuit-under-test of defect will be present
Plate is placed in setting regions.
The defect inspection method of the embodiment of the present application, by obtain circuit-under-test plate shooting image, will shooting image with
Standard picture respective pixel carries out Pixel Information difference, obtains the difference value of respective pixel, wherein standard picture is to not existing
What the reference circuit plate of defect was shot, by shoot each pixel of image Pixel Information and respective pixel difference value into
Row synthesis, obtains input picture, input picture is inputted trained disaggregated model, with according to the output of disaggregated model, really
Determine the defect information of circuit-under-test plate.Since disaggregated model has learnt to obtain the corresponding input picture of circuit board of all kinds of defects
Characteristics of image, therefore, by by the difference value and shooting image Pixel Information synthesized after carry out image characteristics extraction,
The input data that disaggregated model can be enriched helps to improve the precision and efficiency of the detection of circuit-under-test board defect.
The application second aspect embodiment proposes a kind of defect detecting device of circuit board, comprising:
Module is obtained, for obtaining the shooting image of circuit board；
Difference block, for obtaining the shooting image and standard picture respective pixel progress Pixel Information difference pair
Answer the difference value of pixel；Wherein, the standard picture is shot to the reference circuit plate of non-existing defects；
Processing module, for closing the difference value of the Pixel Information of shooting each pixel of image and respective pixel
At obtaining input picture；
Detection module, for the input picture to be inputted trained disaggregated model, according to the disaggregated model
Output, determine the defect information of the circuit-under-test plate；Wherein, the disaggregated model has learnt to obtain the electricity of all kinds of defects
The characteristics of image of the corresponding input picture of road plate.
The defect detecting device of the embodiment of the present application, by the shooting image for obtaining circuit-under-test plate；Will shooting image with
Standard picture respective pixel carries out Pixel Information difference, obtains the difference value of respective pixel, wherein standard picture is to not existing
What the reference circuit plate of defect was shot, by shoot each pixel of image Pixel Information and respective pixel difference value into
Row synthesis, obtains input picture, input picture is inputted trained disaggregated model, with according to the output of disaggregated model, really
Determine the defect information of circuit-under-test plate.Since disaggregated model has learnt to obtain the corresponding input picture of circuit board of all kinds of defects
Characteristics of image, therefore, by by the difference value and shooting image Pixel Information synthesized after carry out image characteristics extraction,
The input data that disaggregated model can be enriched helps to improve the precision and efficiency of the detection of circuit-under-test board defect.
The application third aspect embodiment proposes a kind of computer equipment, comprising: including memory, processor and storage
On a memory and the computer program that can run on a processor, it when the processor executes described program, realizes as above-mentioned
The defect inspection method of circuit board as described in the examples.
The application fourth aspect embodiment proposes a kind of non-transitorycomputer readable storage medium, is stored thereon with meter
Calculation machine program realizes the defect inspection method such as above-mentioned circuit board as described in the examples when the program is executed by processor.
The additional aspect of the application and advantage will be set forth in part in the description, and will partially become from the following description
It obtains obviously, or recognized by the practice of the application.
Detailed description of the invention
The application is above-mentioned and/or additional aspect and advantage will become from the following description of the accompanying drawings of embodiments
Obviously and it is readily appreciated that, in which:
Fig. 1 is a kind of flow diagram of the defect inspection method of circuit board provided by the embodiment of the present application；
Fig. 2 is a kind of flow diagram of disaggregated model training method provided by the embodiment of the present application；
Fig. 3 is a kind of exemplary diagram of disaggregated model training method provided by the embodiment of the present application；
Fig. 4 is the flow diagram of the defect inspection method of another kind circuit board provided by the embodiment of the present application；
Fig. 5 is the flow diagram of the defect inspection method of another circuit board provided by the embodiment of the present application；
Fig. 6 is a kind of structural schematic diagram of the defect detecting device of circuit board provided by the embodiment of the present application；
Fig. 7 shows the block diagram for being suitable for the exemplary computer device for being used to realize the application embodiment.
Specific embodiment
Embodiments herein is described below in detail, examples of the embodiments are shown in the accompanying drawings, wherein from beginning to end
Same or similar label indicates same or similar element or element with the same or similar functions.Below with reference to attached
The embodiment of figure description is exemplary, it is intended to for explaining the application, and should not be understood as the limitation to the application.
The embodiment of the present application relies on pure artificial detection method to the defects detection of circuit board, exists in the prior art
The technical problem that testing cost is high, testing result accuracy rate is lower and real-time is poor, proposes a kind of defect of circuit board
Detection method.
The defect inspection method of circuit board provided by the embodiments of the present application, by obtaining the shooting image of circuit-under-test plate,
Image will be shot and standard picture respective pixel carries out Pixel Information difference, obtain the difference value of respective pixel, wherein standard drawing
Seem to be shot to the reference circuit plate of non-existing defects, the Pixel Information of each pixel of image and corresponding picture will be shot
The difference value of element is synthesized, and obtains input picture, input picture is inputted trained disaggregated model, according to classification mould
The output of type determines the defect information of circuit-under-test plate.
Below with reference to the accompanying drawings the defect inspection method and device of the circuit board of the embodiment of the present application are described.
Fig. 1 is a kind of flow diagram of the defect inspection method of circuit board provided by the embodiment of the present application.
As shown in Figure 1, the defect inspection method of the circuit board the following steps are included:
Step 101, the shooting image of circuit-under-test plate is obtained.
Wherein, circuit-under-test plate refers to the circuit board for needing to carry out defects detection.
In the embodiment of the present application, when carrying out defects detection to circuit board, the shooting of acquisition circuit-under-test plate can be passed through
Image, to carry out defects detection to circuit-under-test plate according to shooting image.Specifically, in the shooting image for obtaining circuit-under-test plate
When, it can use the high-precision camera of image capturing system, by adjusting the angle of camera, light, filter, times mirror, gather
The parameters such as coke, and then the shooting image of circuit-under-test plate is collected in real time.
As an example, if production circuit-under-test plate workshop light than it is darker when, can by adjusting high-precision
The parameters such as angle, the light of camera, and then collect the image of clearly circuit-under-test plate.
Step 102, shooting image and standard picture respective pixel are subjected to Pixel Information difference, obtain the difference of respective pixel
Score value；Wherein, standard picture is shot to the reference circuit plate of non-existing defects.
In the embodiment of the present application, standard picture refer to using high-precision camera to the reference circuit plates of non-existing defects into
The image that row shooting obtains.After getting the shooting image of circuit-under-test plate, each pixel and standard picture of image will be shot
In corresponding pixel carry out the Difference Calculation of Pixel Information, obtain the difference value of respective pixel.
In the embodiment of the present application, the Pixel Information for shooting each pixel in image and standard picture includes depth value and each color
The gray value in channel.The Pixel Information for shooting image and standard picture includes depth value and R, tri- color channels pair of G, B
The gray value answered.
Wherein, depth value refers to the distance between object and camera of respective pixel imaging.The ash of each color channel
Angle value refers to the brightness of each color channel, that is, the depth degree of each color.The range of gray value generally from 0 to
255, white is 255, black 0.
It, can be by the depth value and standard of pixel each in the Pixel Information for shooting image as a kind of possible implementation
The depth value of corresponding pixel carries out Difference Calculation in the Pixel Information of image, obtains the depth difference value of respective pixel.
It, can also be by pixel in the Pixel Information for shooting image in each color channel as alternatively possible implementation
Gray value and pixel corresponding in standard picture carry out Difference Calculation in the gray value of each color channel, obtain respective pixel and exist
The gray scale difference score value of each color channel.
It, can be by pixel in the Pixel Information for shooting image in each color channel as another possible implementation
Gray value of the gray value with pixel corresponding in standard picture in each color channel carries out Difference Calculation, obtains respective pixel each
The gray scale difference score value of color channel；And by the pixel of the depth value of pixel each in the Pixel Information for shooting image and standard picture
The depth value of corresponding pixel carries out Difference Calculation in information, obtains the depth difference value of respective pixel.
Step 103, the difference value of the Pixel Information and respective pixel that shoot image is synthesized, obtains input picture.
It is corresponding with standard picture in each pixel for the shooting image being calculated by pixel difference in the embodiment of the present application
Pixel carry out Pixel Information difference, after obtaining the difference value of respective pixel, the Pixel Information and respective pixel of image will be shot
Difference value synthesized, obtain input picture.
Specifically, shooting image can be obtained with standard picture respective pixel progress Pixel Information Difference Calculation corresponding
The depth difference value of pixel and each color channel gray scale difference score value increase to shooting image in each pixel Pixel Information in,
Obtain the Pixel Information of respective pixel in input picture.At this point, the Pixel Information of each pixel includes the depth of pixel in input picture
Angle value, the gray value of each color channel, depth difference value and the gray scale difference score value in each color channel.
Step 104, input picture is inputted into trained disaggregated model, to determine quilt according to the output of disaggregated model
The defect information of slowdown monitoring circuit plate；Wherein, disaggregated model has learnt to obtain the corresponding input picture of circuit board of all kinds of defects
Characteristics of image.
It should be noted that circuit-under-test plate is limited by reasons such as technique, equipment, raw materials in process of production, it can
Some defects can be generated, therefore, in the embodiment of the present application, whether the defect information of circuit-under-test plate is used to indicate circuit board
Existing defects and defect type.Defect type can weld few tin for chip, chip welds more tin, chip welds even tin, core
The defects of piece setup error.But in the embodiment of the present application, the case where defect information of circuit-under-test plate is likely present remaining,
It is not limited to above-mentioned defect information.
In one embodiment of the application, disaggregated model can pass through depth convolutional neural networks (Deep
Convolutional Neural Network, abbreviation DCNN) training sample image is trained to obtain, and training obtains
Disaggregated model learnt to obtain the characteristics of image of the corresponding input picture of circuit board of all kinds of defects, therefore, by tested electricity
After plate corresponding input picture in road is input to disaggregated model, according to the output of disaggregated model, lacking for circuit-under-test plate can be determined
Fall into information.
Wherein, increased in the Pixel Information for trained sample image sample image and standard picture respective pixel into
The difference value that row Pixel Information difference obtains.
The defect inspection method of the embodiment of the present application, by obtain circuit-under-test plate shooting image, will shooting image with
Standard picture respective pixel carries out Pixel Information difference, obtains the difference value of respective pixel, wherein standard picture is to not existing
What the reference circuit plate of defect was shot, the difference value of the Pixel Information and respective pixel that shoot image is closed
At obtaining input picture, the input picture after synthesis inputted trained disaggregated model, according to the defeated of disaggregated model
Out, the defect information of circuit-under-test plate is determined, wherein the circuit board that disaggregated model has learnt to obtain all kinds of defects is corresponding defeated
Enter the characteristics of image of image.The image for the corresponding input picture of circuit board for obtaining all kinds of defects is learnt due to disaggregated model
Therefore feature carries out image characteristics extraction in the Pixel Information by the way that the difference value to be increased to shooting image, can enrich point
The input data of class model helps to improve the precision and efficiency of the detection of circuit-under-test board defect.
It, can be by sample graph for the training of disaggregated model in a kind of possible way of realization of the embodiment of the present application
It is synthesized as carrying out the difference value that respective pixel difference obtains with standard picture with the Pixel Information of respective pixel in sample image, into
And the sample image after synthesis is used to be trained, referring to fig. 2, Fig. 2 is mentioned specific model training process for the embodiment of the present application
A kind of flow diagram of the disaggregated model training method supplied.
As shown in Fig. 2, the model training method may comprise steps of:
Step 201, sample image and standard picture are carried out in the difference value and sample image that respective pixel difference obtains
The Pixel Information of respective pixel synthesizes.
Wherein, defect information has been marked in sample image, defect information is used to indicate the circuit of respective sample image displaying
Plate whether there is defect and defect type.
In the embodiment of the present application, by the Pixel Information of pixel and standard picture each in the Pixel Information of the sample image of acquisition
Middle respective pixel carries out pixel difference, obtains the difference value of respective pixel.Wherein, sample image is by high-precision camera pair
Sampler plate is shot.
Specifically, the depth value of pixel each in the Pixel Information of sample image is corresponding with the Pixel Information of standard picture
Pixel depth value carry out Difference Calculation, obtain the depth difference value of respective pixel.By picture in the Pixel Information of sample image
Element carries out Difference Calculation in the gray value of each color channel in gray value and the pixel corresponding in standard picture of each color channel,
Respective pixel is obtained in the gray scale difference score value of each color channel.
Further, sample image and standard picture respective pixel are subjected to the difference value and sample that Pixel Information difference obtains
The Pixel Information synthesis of respective pixel in this image, in the sample image after synthesis the Pixel Information of each pixel include depth value,
The gray value and sample image of each color channel and standard picture carry out the depth difference value that respective pixel difference obtains and
The gray scale difference score value of each color channel.
Step 202, it is trained using the sample image after synthesis.
As a kind of possible situation, the disaggregated model in the embodiment of the present application may include convolutional layer, pond layer and complete
Articulamentum.Wherein, convolutional layer carries out image characteristics extraction for the Pixel Information to each pixel in input disaggregated model image；
Pond layer, the feature for extracting to convolutional layer carry out dimensionality reduction operation；Full articulamentum, for according to the image after the layer dimensionality reduction of pond
Feature is classified.
In the embodiment of the present application, using the defect information for the circuit board of sample image marked as the spy of model training
Sign, the sample image after synthesis is input in disaggregated model, is trained to the disaggregated model, and then output circuit plate is scarce
Fall into information.
It is to be understood that using the defect information for the circuit board of sample image marked as the feature of model training,
It is because defect information can clearly represent the circuit board of respective sample image displaying with the presence or absence of defect and defect class
Type inputs the corresponding input picture of circuit-under-test plate to be trained using defect information as feature to disaggregated model
To after disaggregated model, it can determine that circuit-under-test plate whether there is defect according to the defect information of the circuit-under-test plate of output,
And existing defect type.
Specifically, in the sample image input disaggregated model after Pixel Information being synthesized, the convolutional layer of disaggregated model according to
Sample image has marked defect information and has carried out image characteristics extraction to the Pixel Information of each pixel in the sample image of input.For
Image carries out there are two ways to feature extraction, and a kind of method is to be divided automatically to sample image first, marks off sample
Object or color region included in image then according to these extracted region images features, and establish index；Another kind side
Sample image is simply evenly divided into several regular sub-blocks by rule, then extracts feature to each sample image sub-block,
And establish index.
Further, pond layer carries out dimensionality reduction operation to the characteristics of image that convolutional layer extracts, and only retains in characteristics of image
Main feature classifies to the characteristics of image after the layer dimensionality reduction of pond by full articulamentum, thus after disaggregated model output category
Defect information.
Step 203, when the defect information of disaggregated model output and the defect information difference of sample image mark minimize, point
Class model training terminates.
In the embodiment of the present application, when disaggregated model output defect information and sample image mark defect information difference compared with
When big, further continue to be trained disaggregated model by the way of signal propagated forward, error back propagation, until classification
When the defect information of model output and the defect information difference of sample image mark minimize, terminate disaggregated model training.
As a kind of possible situation, when the defect information that defect information and the sample image of disaggregated model output mark is poor
It is different when reaching preset discrepancy threshold, terminate the training to disaggregated model.
Further, the mode that the disaggregated model after training is online by small flow, gradually replaces and just runs on line
Disaggregated model, so that it is extensive to achieve the purpose that disaggregated model is extended with service dynamic.
As an example, referring to Fig. 3, Fig. 3 is a kind of according to disaggregated model output quilt for what is provided in the embodiment of the present application
The exemplary diagram of the defect information of slowdown monitoring circuit plate.Disaggregated model in Fig. 3 includes convolutional layer, pond layer and full articulamentum.Passing through will
The shooting image and standard picture respective pixel of circuit-under-test plate carry out the difference value that Pixel Information difference obtains and shooting image
Pixel Information synthesized, after determining input picture, input picture is inputted into disaggregated model shown in Fig. 3, this point
The convolutional layer of class model carries out image characteristics extraction to the Pixel Information of pixel each in input picture, and further, pond layer is logical
The method for crossing image down sampling carries out dimensionality reduction operation to the characteristics of image extracted, obtains the primary picture feature of input picture,
Classify eventually by full articulamentum to the characteristics of image after dimensionality reduction, to determine tested electricity according to the defect information of output
Road plate is with the presence or absence of defect and the defect type of existing defects.
It should be noted that the characteristics of for different board production scenes and data, can be based on shown in Fig. 3 point
Class model, the deep neural network mould that design different depth, different number neuron, different convolution pond organizational forms are constituted
Type, then using the sample image for having marked defect information, to model by the way of signal propagated forward, error back propagation
It is trained, to meet the needs of different business scene.
In the embodiment of the present application, by the way that sample image and standard picture respective pixel are carried out what Pixel Information difference obtained
Difference value and the Pixel Information of respective pixel in sample image are synthesized, and the sample image of the Pixel Information after synthesis is carried out
Training, when the defect information difference of the defect information of disaggregated model output and sample image mark minimizes, disaggregated model instruction
White silk terminates.As a result, by the training to disaggregated model, can be realized after input picture is inputted trained disaggregated model,
It can accurately determine that circuit-under-test plate whether there is defect information and defect type according to the output of disaggregated model, thus
Improve the accuracy of circuit board detecting.
In the embodiment of the present application, on the basis of the embodiment described in Fig. 1, according to the output of disaggregated model, tested electricity is determined
After the defect information of road plate, further judge the circuit-under-test plate with the presence or absence of scarce according to the defect information of circuit-under-test plate
It falls into.
As a kind of possible situation, defect information existing for the circuit-under-test plate exported according to disaggregated model determines quilt
After slowdown monitoring circuit plate existing defects, lacked according to what the shooting image and artificial defect type of the circuit board of existing defects marked
Sunken information obtains the first training sample, and then is trained using the first training sample to disaggregated model.It is carried out below with reference to Fig. 4
It is described in detail, Fig. 4 is the flow diagram of the defect inspection method of another circuit board provided by the embodiments of the present application.
As shown in figure 4, further comprising the steps of after the step 104 of previous embodiment:
Step 301, if circuit-under-test plate existing defects, shooting image is shown in control interface.
It is in the embodiment of the present application, the shooting image of circuit-under-test plate and standard picture respective pixel progress Pixel Information is poor
The Pixel Information of the difference value got and shooting image is synthesized, and determines input picture, and then input picture is inputted
Disaggregated model determines the defect information of circuit-under-test plate, since the defect information of circuit-under-test plate can determine shooting image exhibition
The circuit board shown whether there is defect and existing defect type.When according to the defect information of circuit-under-test plate determine by
When slowdown monitoring circuit plate existing defects, the shooting image of circuit-under-test plate is shown in control interface.
In the embodiment of the present application, when determining circuit-under-test plate existing defects according to the defect information of circuit-under-test plate,
It can be placed in setting regions by controlling production line with the circuit-under-test plate of defect will be present.
For example, scarce when determining that circuit-under-test plate just in production exists according to the defect information of circuit-under-test plate
When falling into, it can control mechanical arm and take out the circuit board from production line, conveyer belt can also be controlled be sent to the circuit board and set
Positioning is set, in order to separate normal circuit plate and abnormal circuit plate, with the circuit board centralized processing to existing defects.It separates herein
Only as an example, specific separate mode determines the mode of abnormal circuit plate according to actual production line.
When separating normal circuit plate and abnormal circuit plate, difference can be preset according to the defect type of circuit board
Region, be placed in different regions so that the circuit board of different defect types will be present, in order to use different processing modes pair
There are the circuit boards of different defect types to be handled.The circuit board that defect can also will be present, which is placed in the same area, to be collected
Middle processing.Specifically processing mode according to the actual situation depending on, this is not construed as limiting in the embodiment of the present application.
As alternatively possible situation, lacked when determining that circuit-under-test plate exists according to the defect information of circuit-under-test plate
When falling into, can set out preset alarm system, be placed in setting regions so that the circuit-under-test plate of defect will be present, can also will be by
The defect information of slowdown monitoring circuit plate is stored into Production database, to be updated to Production database, thus according to raw after update
The data produced in database are trained disaggregated model.
Step 302, it obtains and the defect information that artificial defect type marks is carried out to shooting image.
In the embodiment of the present application, when circuit-under-test plate existing defects, control interface show shooting image after, pass through by
It shoots image and standard picture carries out comparison in difference, can get and what artificial defect type marked is carried out to shooting image
Defect information.
As an example it is assumed that after carrying out artificial defect type mark to circuit-under-test plate, defective type A, defect type
B, defect type C.It is defect class that defect type existing for the circuit board, which can be got, by the shooting image of circuit-under-test plate
Type B.
Step 303, according to shooting image and defect information, the first training sample is generated.
In the embodiment of the present application, after in shooting image and the first training sample of defect information generation, the training sample
This sample image has been labelled with defect information, i.e., the training sample existing defects and defect have been marked in sample image
Type.
Step 304, disaggregated model is trained using the first training sample.
In the embodiment of the present application, above-mentioned reality may refer to the method that disaggregated model is trained using the first training sample
Disaggregated model training method described in example is applied, details are not described herein.
In the embodiment of the present application, when determining circuit-under-test plate existing defects by the defect information of circuit-under-test, controlling
Showing interface shoots image, obtains and carries out the defect information that artificial defect type marks to shooting image, according to shooting figure
Picture and defect information are generated the first training sample, are trained using the first training sample to disaggregated model.Basis is deposited as a result,
Training sample is obtained in the shooting image and defect information of the circuit board of defect, to be trained to disaggregated model, can be realized
Update to disaggregated model, to improve the precision of circuit board detecting.
In order to avoid the circuit board to existing defects carries out overlearning, lead to model training deviation, in Fig. 4 embodiment
On the basis of, can also further according to disaggregated model export circuit-under-test plate existing for defect information, determine circuit-under-test plate
After non-existing defects, according to the defect information manually marked in the shooting image and image of partial circuit plate, the second instruction is generated
Practice sample, and then disaggregated model is trained using the first training sample and the second training sample.It is carried out below with reference to Fig. 5 detailed
Describe in detail it is bright, Fig. 5 be another circuit board provided by the embodiments of the present application defect inspection method flow diagram.
As shown in figure 5, further comprising the steps of after the step 104 of previous embodiment:
Step 401, never selected part circuit board in the circuit-under-test plate of existing defects shows in control interface and chooses
The shooting image of partial circuit plate.
In the embodiment of the present application, corresponding circuit board can be determined with the presence or absence of scarce according to the defect information of circuit-under-test plate
It falls into, the selected part circuit board from the circuit-under-test plate there is no defect having determined, further, shows and select in control interface
The shooting image of the partial circuit plate taken.
It should be noted that the quantity due to the circuit-under-test plate that defect is not present in production process can be very much, if by
One there will be no defect circuit-under-test plate shooting image control interface show and artificial detection will be one it is huge
Engineering, therefore it may only be necessary to selected part there is no defect circuit board show, manually to recheck.
Step 402, to the partial circuit plate of selection, the defect information that corresponding shooting image manually marks is obtained, to generate
The second training sample for being trained to disaggregated model.
In the embodiment of the present application, to the partial circuit chosen in the circuit board of never existing defects by artificial reinspection confirmation
Plate marks corresponding defect information, i.e., there is no the marks of defect for mark, and then generate the second training sample, using generation
The second training sample disaggregated model is trained.
It should be noted that above-mentioned reality can also be referred to the process that disaggregated model is trained using the second training sample
It applies to the method for disaggregated model training in example, details are not described herein.
In the embodiment of the present application, by selected part circuit board in the circuit-under-test plate of never existing defects, on control circle
Face shows the shooting image for the partial circuit plate chosen, and to the partial circuit plate of selection, obtains corresponding shooting image and manually marks
Defect information, to generate the second training sample for being trained to disaggregated model.As a result, by the way that there is no the electricity of defect
The training sample that the corresponding shooting image of road plate and defect information generate is trained disaggregated model, can be realized to classification mould
The real-time update of type, to improve the precision of circuit board detecting.
In order to realize above-described embodiment, the application also proposes a kind of defect detecting device of circuit board.
Fig. 6 is a kind of structural schematic diagram of the defect detecting device of circuit board provided by the embodiments of the present application.
As shown in fig. 6, the defect detecting device 100 of the circuit board includes: to obtain module 110, difference block 120, processing
Module 130 and detection module 140.
Module 110 is obtained, for obtaining the shooting image of circuit board.
Difference block 120 carries out Pixel Information difference for that will shoot image and standard picture respective pixel, is corresponded to
The difference value of pixel；Wherein, standard picture is shot to the reference circuit plate of non-existing defects.
Processing module 130, for the difference value of the Pixel Information and respective pixel that shoot each pixel of image to be synthesized,
Obtain input picture.
Detection module 140, for input picture to be inputted trained disaggregated model, according to the defeated of disaggregated model
Out, the defect information of circuit-under-test plate is determined；Wherein, disaggregated model learnt to obtain all kinds of defects circuit board it is corresponding defeated
Enter the characteristics of image of image.
As a kind of possible situation, shoot each pixel in image and standard picture Pixel Information include depth value and
The gray value of each color channel.
As alternatively possible situation, difference block 120 is specifically used for:
Respective pixel in image and standard picture will be shot and carry out depth value difference, obtain the depth difference of respective pixel
Value；Respective pixel in image and standard picture will be shot and carry out gray value difference, obtain respective pixel in the ash of each color channel
Spend difference value.
As alternatively possible situation, processing module 130 is specifically used for: by depth difference value, depth value, and it is each
Pixel Information of the gray scale difference score value and gray value of color channel as respective pixel in input picture.
As alternatively possible situation, disaggregated model includes convolutional layer, pond layer and full articulamentum；
Wherein, convolutional layer carries out image characteristics extraction for the Pixel Information to pixel each in input picture；
Pond layer, the feature for extracting to convolutional layer carry out dimensionality reduction operation；
Full articulamentum, for being classified according to the characteristics of image after the layer dimensionality reduction of pond.
As alternatively possible situation, disaggregated model is that sample image and standard picture respective pixel are carried out pixel
The difference value that information gap is got is synthesized with the Pixel Information of respective pixel in sample image, using the sample image after synthesis into
Row training obtains；Wherein, sample image has marked defect information, and defect information is used to indicate the electricity of respective sample image displaying
Road plate whether there is defect and defect type；
When the defect information difference of the defect information of disaggregated model output and sample image mark minimizes, disaggregated model
Training terminates.
As alternatively possible situation, defect detecting device 100, further includes:
First display module shows shooting image in control interface if being used for circuit-under-test plate existing defects.
Second obtains module, carries out the defect information that artificial defect type marks to shooting image for obtaining.
First generation module, for generating the first training sample according to shooting image and defect information.
Training module, for being trained using the first training sample to disaggregated model.
As alternatively possible situation, defect detecting device 100, further includes:
Second display module, for selected part circuit board in the circuit-under-test plate of never existing defects, in control interface
Show the shooting image for the partial circuit plate chosen.
Second generation module obtains the defect letter that corresponding shooting image manually marks for the partial circuit plate to selection
Breath, to generate the second training sample for being trained to disaggregated model.
As alternatively possible situation, defect detecting device 100, further includes:
Control module controls production line if being used for circuit-under-test plate existing defects, the tested of defect will be present
Circuit board is placed in setting regions.
It should be noted that the aforementioned defect for being also applied for the embodiment to the explanation of defect inspection method embodiment
Detection device, details are not described herein again.
The defect detecting device of the embodiment of the present application, by the shooting image for obtaining circuit-under-test plate；Will shooting image with
Standard picture respective pixel carries out Pixel Information difference, obtains the difference value of respective pixel, wherein standard picture is to not existing
What the reference circuit plate of defect was shot, by shoot each pixel of image Pixel Information and respective pixel difference value into
Row synthesis, obtains input picture, input picture is inputted trained disaggregated model, with according to the output of disaggregated model, really
Determine the defect information of circuit-under-test plate, wherein disaggregated model has learnt to obtain the corresponding input figure of circuit board of all kinds of defects
The characteristics of image of picture.The image for having learnt the corresponding input picture of circuit board for obtaining all kinds of defects due to disaggregated model is special
Therefore sign carries out image characteristics extraction in the Pixel Information by the way that the difference value to be increased to shooting image, can enrich classification
The input data of model helps to improve the precision and efficiency of the detection of circuit-under-test board defect.
In order to realize above-described embodiment, the application also proposes a kind of computer equipment, comprising: including memory, processor
And the computer program that can be run on a memory and on a processor is stored, when the processor executes described program, realize
Such as the defect inspection method of above-mentioned circuit board as described in the examples.
In order to realize above-described embodiment, the application also proposes a kind of non-transitorycomputer readable storage medium, deposits thereon
Computer program is contained, the defects detection side such as above-mentioned circuit board as described in the examples is realized when which is executed by processor
Method.
Fig. 7 shows the block diagram for being suitable for the exemplary computer device for being used to realize the application embodiment.What Fig. 7 was shown
Computer equipment 12 is only an example, should not function to the embodiment of the present application and use scope bring any restrictions.
As shown in fig. 7, computer equipment 12 is showed in the form of universal computing device.The component of computer equipment 12 can be with
Including but not limited to: one or more processor or processing unit 16, system storage 28 connect different system components
The bus 18 of (including system storage 28 and processing unit 16).
Bus 18 indicates one of a few class bus structures or a variety of, including memory bus or Memory Controller,
Peripheral bus, graphics acceleration port, processor or the local bus using any bus structures in a variety of bus structures.It lifts
For example, these architectures include but is not limited to industry standard architecture (Industry Standard
Architecture；Hereinafter referred to as: ISA) bus, microchannel architecture (Micro Channel Architecture；Below
Referred to as: MAC) bus, enhanced isa bus, Video Electronics Standards Association (Video Electronics Standards
Association；Hereinafter referred to as: VESA) local bus and peripheral component interconnection (Peripheral Component
Interconnection；Hereinafter referred to as: PCI) bus.
Computer equipment 12 typically comprises a variety of computer system readable media.These media can be it is any can be by
The usable medium that computer equipment 12 accesses, including volatile and non-volatile media, moveable and immovable medium.
Memory 28 may include the computer system readable media of form of volatile memory, such as random access memory
Device (Random Access Memory；Hereinafter referred to as: RAM) 30 and/or cache memory 32.Computer equipment 12 can be with
It further comprise other removable/nonremovable, volatile/non-volatile computer system storage mediums.Only as an example,
Storage system 34 can be used for reading and writing immovable, non-volatile magnetic media, and (Fig. 7 do not show, commonly referred to as " hard drive
Device ").Although being not shown in Fig. 7, the disk for reading and writing to removable non-volatile magnetic disk (such as " floppy disk ") can be provided and driven
Dynamic device, and to removable anonvolatile optical disk (such as: compact disc read-only memory (Compact Disc Read Only
Memory；Hereinafter referred to as: CD-ROM), digital multi CD-ROM (Digital Video Disc Read Only
Memory；Hereinafter referred to as: DVD-ROM) or other optical mediums) read-write CD drive.In these cases, each driving
Device can be connected by one or more data media interfaces with bus 18.Memory 28 may include that at least one program produces
Product, the program product have one group of (for example, at least one) program module, and it is each that these program modules are configured to perform the application
The function of embodiment.
Program/utility 40 with one group of (at least one) program module 42 can store in such as memory 28
In, such program module 42 include but is not limited to operating system, one or more application program, other program modules and
It may include the realization of network environment in program data, each of these examples or certain combination.Program module 42 is usual
Execute the function and/or method in embodiments described herein.
Computer equipment 12 can also be with one or more external equipments 14 (such as keyboard, sensing equipment, display 24
Deng) communication, the equipment interacted with the computer system/server 12 can be also enabled a user to one or more to be communicated, and/
Or with the computer equipment 12 is communicated with one or more of the other calculating equipment any equipment (such as network interface card,
Modem etc.) communication.This communication can be carried out by input/output (I/O) interface 22.Also, computer equipment
12 can also pass through network adapter 20 and one or more network (such as local area network (Local Area Network；Below
Referred to as: LAN), wide area network (Wide Area Network；Hereinafter referred to as: WAN) and/or public network, for example, internet) it is logical
Letter.As shown in fig. 7, network adapter 20 is communicated by bus 18 with other modules of computer equipment 12.Although should be understood that
It is not shown in the figure, other hardware and/or software module can be used in conjunction with computer equipment 12, including but not limited to: microcode,
Device driver, redundant processing unit, external disk drive array, RAID system, tape drive and data backup storage
System etc..
Processing unit 16 by the program that is stored in system storage 28 of operation, thereby executing various function application and
Data processing, such as realize the defect inspection method of the circuit board referred in previous embodiment.
In the description of this specification, reference term " one embodiment ", " some embodiments ", " example ", " specifically show
The description of example " or " some examples " etc. means specific features, structure, material or spy described in conjunction with this embodiment or example
Point is contained at least one embodiment or example of the application.In the present specification, schematic expression of the above terms are not
It must be directed to identical embodiment or example.Moreover, particular features, structures, materials, or characteristics described can be in office
It can be combined in any suitable manner in one or more embodiment or examples.In addition, without conflicting with each other, the skill of this field
Art personnel can tie the feature of different embodiments or examples described in this specification and different embodiments or examples
It closes and combines.
In addition, term " first ", " second " are used for descriptive purposes only and cannot be understood as indicating or suggesting relative importance
Or implicitly indicate the quantity of indicated technical characteristic.Define " first " as a result, the feature of " second " can be expressed or
Implicitly include at least one this feature.In the description of the present application, the meaning of " plurality " is at least two, such as two, three
It is a etc., unless otherwise specifically defined.
Any process described otherwise above or method description are construed as in flow chart or herein, and expression includes
It is one or more for realizing custom logic function or process the step of executable instruction code module, segment or portion
Point, and the range of the preferred embodiment of the application includes other realization, wherein can not press shown or discussed suitable
Sequence, including according to related function by it is basic simultaneously in the way of or in the opposite order, Lai Zhihang function, this should be by the application
Embodiment person of ordinary skill in the field understood.
Expression or logic and/or step described otherwise above herein in flow charts, for example, being considered use
In the order list for the executable instruction for realizing logic function, may be embodied in any computer-readable medium, for
Instruction execution system, device or equipment (such as computer based system, including the system of processor or other can be held from instruction
The instruction fetch of row system, device or equipment and the system executed instruction) it uses, or combine these instruction execution systems, device or set
It is standby and use.For the purpose of this specification, " computer-readable medium ", which can be, any may include, stores, communicates, propagates or pass
Defeated program is for instruction execution system, device or equipment or the dress used in conjunction with these instruction execution systems, device or equipment
It sets.The more specific example (non-exhaustive list) of computer-readable medium include the following: there is the electricity of one or more wirings
Interconnecting piece (electronic device), portable computer diskette box (magnetic device), random access memory (RAM), read-only memory
(ROM), erasable edit read-only storage (EPROM or flash memory), fiber device and portable optic disk is read-only deposits
Reservoir (CDROM).In addition, computer-readable medium can even is that the paper that can print described program on it or other are suitable
Medium, because can then be edited, be interpreted or when necessary with it for example by carrying out optical scanner to paper or other media
His suitable method is handled electronically to obtain described program, is then stored in computer storage.
It should be appreciated that each section of the application can be realized with hardware, software, firmware or their combination.Above-mentioned
In embodiment, software that multiple steps or method can be executed in memory and by suitable instruction execution system with storage
Or firmware is realized.Such as, if realized with hardware in another embodiment, following skill well known in the art can be used
Any one of art or their combination are realized: have for data-signal is realized the logic gates of logic function from
Logic circuit is dissipated, the specific integrated circuit with suitable combinational logic gate circuit, programmable gate array (PGA), scene can compile
Journey gate array (FPGA) etc..
Those skilled in the art are understood that realize all or part of step that above-described embodiment method carries
It suddenly is that relevant hardware can be instructed to complete by program, the program can store in a kind of computer-readable storage medium
In matter, which when being executed, includes the steps that one or a combination set of embodiment of the method.
It, can also be in addition, can integrate in a processing module in each functional unit in each embodiment of the application
It is that each unit physically exists alone, can also be integrated in two or more units in a module.Above-mentioned integrated mould
Block both can take the form of hardware realization, can also be realized in the form of software function module.The integrated module is such as
Fruit is realized and when sold or used as an independent product in the form of software function module, also can store in a computer
In read/write memory medium.
Storage medium mentioned above can be read-only memory, disk or CD etc..Although having been shown and retouching above
Embodiments herein is stated, it is to be understood that above-described embodiment is exemplary, and should not be understood as the limit to the application
System, those skilled in the art can be changed above-described embodiment, modify, replace and become within the scope of application
Type.
Claims (20)
1. a kind of defect inspection method of circuit board, which is characterized in that the described method comprises the following steps:
Obtain the shooting image of circuit-under-test plate；
The shooting image and standard picture respective pixel are subjected to Pixel Information difference, obtain the difference value of respective pixel；Its
In, the standard picture is shot to the reference circuit plate of non-existing defects；
The difference value of the Pixel Information of shooting each pixel of image and respective pixel is synthesized, input picture is obtained；
The input picture is inputted into trained disaggregated model, to determine the quilt according to the output of the disaggregated model
The defect information of slowdown monitoring circuit plate；Wherein, the disaggregated model has learnt to obtain the corresponding input figure of circuit board of all kinds of defects
The characteristics of image of picture.
2. defect inspection method according to claim 1, which is characterized in that in the shooting image and the standard picture
The Pixel Information of each pixel includes the gray value of depth value and each color channel.
3. defect inspection method according to claim 2, which is characterized in that described by the shooting image and standard picture
Respective pixel carries out Pixel Information difference, obtains the difference value of respective pixel, comprising:
Respective pixel in the shooting image and the standard picture is subjected to depth value difference, obtains the depth difference of respective pixel
Score value；
Respective pixel in the shooting image and the standard picture is subjected to gray value difference, obtains respective pixel in each color
The gray scale difference score value in channel.
4. defect inspection method according to claim 3, which is characterized in that the picture by shooting each pixel of image
The difference value of prime information and respective pixel is synthesized, and input picture is obtained, comprising:
Using the gray scale difference score value of the depth difference value, the depth value and each color channel and gray value as described defeated
Enter the Pixel Information of respective pixel in image.
5. defect inspection method according to claim 1-4, which is characterized in that the disaggregated model includes convolution
Layer, pond layer and full articulamentum；
Wherein, the convolutional layer carries out image characteristics extraction for the Pixel Information to each pixel in the input picture；
The pond layer, the feature for extracting to the convolutional layer carry out dimensionality reduction operation；
The full articulamentum, for being classified according to the characteristics of image after the pond layer dimensionality reduction.
6. defect inspection method according to claim 1-4, which is characterized in that the disaggregated model is by sample
The difference value that this image is obtained with standard picture respective pixel progress Pixel Information difference is corresponding with the sample image
The Pixel Information of pixel synthesizes, and is trained using the sample image after synthesis；Wherein, the sample image has marked
Defect information, the defect information are used to indicate the circuit board of respective sample image displaying with the presence or absence of defect and defect class
Type；
It is described when the defect information difference of the defect information of disaggregated model output and sample image mark minimizes
Disaggregated model training terminates.
7. defect inspection method according to claim 1-4, which is characterized in that described that the input picture is defeated
Enter trained disaggregated model, with according to the output of the disaggregated model, determine the circuit-under-test plate defect information it
Afterwards, further includes:
If the circuit-under-test plate existing defects, the shooting image is shown in control interface；
It obtains and the defect information that artificial defect type marks is carried out to the shooting image；
According to the shooting image and the defect information, the first training sample is generated；
The disaggregated model is trained using first training sample.
8. defect inspection method according to claim 1-4, which is characterized in that described that the input picture is defeated
Enter trained disaggregated model, with according to the output of the disaggregated model, determine the circuit-under-test plate defect information it
Afterwards, further includes:
Never selected part circuit board in the circuit-under-test plate of existing defects shows the partial circuit plate chosen in control interface
Shoot image；
To the partial circuit plate of selection, the defect information that corresponding shooting image manually marks is obtained, to generate for described point
The second training sample that class model is trained.
9. defect inspection method according to claim 1-4, which is characterized in that described that the input picture is defeated
Enter trained disaggregated model, with according to the output of the disaggregated model, determine the circuit-under-test plate defect information it
Afterwards, further includes:
If the circuit-under-test plate existing defects, control production line, set so that the circuit-under-test plate of defect will be present
In in setting regions.
10. a kind of defect detecting device of circuit board, which is characterized in that described device includes:
Module is obtained, for obtaining the shooting image of circuit board；
Difference block obtains corresponding picture for the shooting image and standard picture respective pixel to be carried out Pixel Information difference
The difference value of element；Wherein, the standard picture is shot to the reference circuit plate of non-existing defects；
Processing module is obtained for synthesizing the difference value of the Pixel Information of shooting each pixel of image and respective pixel
To input picture；
Detection module, for the input picture to be inputted trained disaggregated model, according to the defeated of the disaggregated model
Out, the defect information of the circuit-under-test plate is determined；Wherein, the disaggregated model has learnt to obtain the circuit board of all kinds of defects
The characteristics of image of corresponding input picture.
11. defect detecting device according to claim 10, which is characterized in that the shooting image and the standard picture
In each pixel Pixel Information include depth value and each color channel gray value.
12. defect detecting device according to claim 11, which is characterized in that the difference block is specifically used for:
Respective pixel in the shooting image and the standard picture is subjected to depth value difference, obtains the depth difference of respective pixel
Score value；
Respective pixel in the shooting image and the standard picture is subjected to gray value difference, obtains respective pixel in each color
The gray scale difference score value in channel.
13. defect detecting device according to claim 12, which is characterized in that the processing module is specifically used for: by institute
The gray scale difference score value and gray value for stating depth difference value, the depth value and each color channel are as in the input picture
The Pixel Information of respective pixel.
14. the described in any item defect detecting devices of 0-13 according to claim 1, which is characterized in that the disaggregated model includes
Convolutional layer, pond layer and full articulamentum；
Wherein, the convolutional layer carries out image characteristics extraction for the Pixel Information to each pixel in the input picture；
The pond layer, the feature for extracting to the convolutional layer carry out dimensionality reduction operation；
The full articulamentum, for being classified according to the characteristics of image after the pond layer dimensionality reduction.
15. the described in any item defect detecting devices of 0-13 according to claim 1, which is characterized in that the disaggregated model, be by
Sample image and the standard picture respective pixel carry out difference value that Pixel Information difference obtains with it is right in the sample image
It answers the Pixel Information of pixel to synthesize, is trained using the sample image after synthesis；Wherein, the sample image has been marked
Defect information is infused, the defect information is used to indicate the circuit board of respective sample image displaying with the presence or absence of defect and defect
Type；
It is described when the defect information difference of the defect information of disaggregated model output and sample image mark minimizes
Disaggregated model training terminates.
16. the described in any item defect detecting devices of 0-15 according to claim 1, which is characterized in that the defect detecting device,
Further include:
First display module shows the shooting image in control interface if being used for the circuit-under-test plate existing defects；
Second obtains module, carries out the defect information that artificial defect type marks to the shooting image for obtaining；
First generation module, for generating the first training sample according to the shooting image and the defect information；
Training module, for being trained using first training sample to the disaggregated model.
17. the described in any item defect detecting devices of 0-15 according to claim 1, which is characterized in that the defect detecting device,
Further include:
Second display module is shown for selected part circuit board in the circuit-under-test plate of never existing defects in control interface
The shooting image of the partial circuit plate of selection；
Second generation module obtains the defect information that corresponding shooting image manually marks for the partial circuit plate to selection, with
Generate the second training sample for being trained to the disaggregated model.
18. the described in any item defect detecting devices of 0-15 according to claim 1, which is characterized in that the defect detecting device,
Further include:
Control module controls production line, if being used for the circuit-under-test plate existing defects will be present described in defect
Circuit-under-test plate is placed in setting regions.
19. a kind of computer equipment, which is characterized in that including memory, processor and store on a memory and can handle
The computer program run on device when the processor executes described program, realizes the electricity as described in any in claim 1-9
The defect inspection method of road plate.
20. a kind of non-transitorycomputer readable storage medium, is stored thereon with computer program, which is characterized in that the program
The defect inspection method of the circuit board as described in any in claim 1-9 is realized when being executed by processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910133359.9A CN109871895B (en) | 2019-02-22 | 2019-02-22 | Method and device for detecting defects of circuit board |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910133359.9A CN109871895B (en) | 2019-02-22 | 2019-02-22 | Method and device for detecting defects of circuit board |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109871895A true CN109871895A (en) | 2019-06-11 |
| CN109871895B CN109871895B (en) | 2021-03-16 |
Family
ID=66919116
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910133359.9A Active CN109871895B (en) | 2019-02-22 | 2019-02-22 | Method and device for detecting defects of circuit board |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109871895B (en) |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110441315A (en) * | 2019-08-02 | 2019-11-12 | 英特尔产品（成都）有限公司 | Electronic component test equipment and method |
| CN111078912A (en) * | 2019-12-18 | 2020-04-28 | 国网上海市电力公司 | Power equipment image data warehouse and power equipment defect detection method |
| CN111696092A (en) * | 2020-06-11 | 2020-09-22 | 深圳市华汉伟业科技有限公司 | Defect detection method and system based on feature comparison and storage medium |
| CN111986178A (en) * | 2020-08-21 | 2020-11-24 | 北京百度网讯科技有限公司 | Product defect detection method and device, electronic equipment and storage medium |
| CN112026350A (en) * | 2020-09-17 | 2020-12-04 | 广州夕千科技有限公司 | Printing machine for processing electronic components |
| CN112396599A (en) * | 2020-12-04 | 2021-02-23 | 惠州高视科技有限公司 | Visual inspection method for transparent packaged IC (integrated circuit) defects |
| CN112834518A (en) * | 2021-01-06 | 2021-05-25 | 优刻得科技股份有限公司 | Particle defect detection method, system, device and medium |
| CN113111828A (en) * | 2021-04-23 | 2021-07-13 | 中国科学院宁波材料技术与工程研究所 | Three-dimensional defect detection method and system for bearing |
| 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 |
| CN113466261A (en) * | 2021-07-26 | 2021-10-01 | 鸿安（福建）机械有限公司 | PCB board automatic checkout device |
| CN114152627A (en) * | 2022-02-09 | 2022-03-08 | 季华实验室 | Chip circuit defect detection method and device, electronic equipment and storage medium |
| CN114581362A (en) * | 2021-07-22 | 2022-06-03 | 正泰集团研发中心(上海)有限公司 | Photovoltaic module defect detection method and device, electronic equipment and readable storage medium |
| CN114708266A (en) * | 2022-06-07 | 2022-07-05 | 青岛通产智能科技股份有限公司 | Tool, method and device for detecting card defects and medium |
| CN115239734A (en) * | 2022-09-23 | 2022-10-25 | 成都数之联科技股份有限公司 | Model training method, device, storage medium, equipment and computer program product |
| TWI803393B (en) * | 2022-07-19 | 2023-05-21 | 神通資訊科技股份有限公司 | Defect detection system based on unsupervised learning and method thereof |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20010033683A1 (en) * | 2000-04-25 | 2001-10-25 | Maki Tanaka | Method of inspecting a pattern and an apparatus thereof and a method of processing a specimen |
| US20030001598A1 (en) * | 2001-06-29 | 2003-01-02 | Kla-Tencor Corporation | Apparatus and methods for reliable and efficient detection of voltage contrast defects |
| CN101339607A (en) * | 2008-08-15 | 2009-01-07 | 北京中星微电子有限公司 | Human face recognition method and system, human face recognition model training method and system |
| CN104458755A (en) * | 2014-11-26 | 2015-03-25 | 吴晓军 | Multi-type material surface defect detection method based on machine vision |
| CN106204618A (en) * | 2016-07-20 | 2016-12-07 | 南京文采科技有限责任公司 | Product surface of package defects detection based on machine vision and sorting technique |
| CN106570511A (en) * | 2016-11-02 | 2017-04-19 | 上海大学 | Keyboard character defect online detection system based on computer vision and keyboard character defect online detection method thereof |
| CN106897994A (en) * | 2017-01-20 | 2017-06-27 | 北京京仪仪器仪表研究总院有限公司 | A kind of pcb board defect detecting system and method based on layered image |
| CN107966448A (en) * | 2017-11-17 | 2018-04-27 | 福建工程学院 | A kind of 2 dimension detection methods for PCB paste solder printing quality |
| CN108982544A (en) * | 2018-06-20 | 2018-12-11 | 青岛联合创智科技有限公司 | A kind of printed circuit board flaw component detection method |
| CN109300151A (en) * | 2018-07-02 | 2019-02-01 | 浙江商汤科技开发有限公司 | Image processing method and device, electronic equipment |
| CN109345524A (en) * | 2018-09-26 | 2019-02-15 | 深圳市鑫汇达机械设计有限公司 | A kind of bearing open defect detection system of view-based access control model |
-
2019
- 2019-02-22 CN CN201910133359.9A patent/CN109871895B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20010033683A1 (en) * | 2000-04-25 | 2001-10-25 | Maki Tanaka | Method of inspecting a pattern and an apparatus thereof and a method of processing a specimen |
| US20030001598A1 (en) * | 2001-06-29 | 2003-01-02 | Kla-Tencor Corporation | Apparatus and methods for reliable and efficient detection of voltage contrast defects |
| CN101339607A (en) * | 2008-08-15 | 2009-01-07 | 北京中星微电子有限公司 | Human face recognition method and system, human face recognition model training method and system |
| CN104458755A (en) * | 2014-11-26 | 2015-03-25 | 吴晓军 | Multi-type material surface defect detection method based on machine vision |
| CN106204618A (en) * | 2016-07-20 | 2016-12-07 | 南京文采科技有限责任公司 | Product surface of package defects detection based on machine vision and sorting technique |
| CN106570511A (en) * | 2016-11-02 | 2017-04-19 | 上海大学 | Keyboard character defect online detection system based on computer vision and keyboard character defect online detection method thereof |
| CN106897994A (en) * | 2017-01-20 | 2017-06-27 | 北京京仪仪器仪表研究总院有限公司 | A kind of pcb board defect detecting system and method based on layered image |
| CN107966448A (en) * | 2017-11-17 | 2018-04-27 | 福建工程学院 | A kind of 2 dimension detection methods for PCB paste solder printing quality |
| CN108982544A (en) * | 2018-06-20 | 2018-12-11 | 青岛联合创智科技有限公司 | A kind of printed circuit board flaw component detection method |
| CN109300151A (en) * | 2018-07-02 | 2019-02-01 | 浙江商汤科技开发有限公司 | Image processing method and device, electronic equipment |
| CN109345524A (en) * | 2018-09-26 | 2019-02-15 | 深圳市鑫汇达机械设计有限公司 | A kind of bearing open defect detection system of view-based access control model |
Non-Patent Citations (3)
| Title |
|---|
| FENG XIE ET AL.: "Detecting PCB component placement defects by genetic programming", 《2013 IEEE CONGRESS ON EVOLUTIONARY COMPUTATION》 * |
| SANLI TANG ET AL.: "ONLINE PCB DEFECT DETECTOR ON A NEW PCB DEFECT DATASET", 《ARXIV》 * |
| 王栋等: "基于形态学的PCB缺陷快速检测技术", 《计算机科学》 * |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110441315A (en) * | 2019-08-02 | 2019-11-12 | 英特尔产品（成都）有限公司 | Electronic component test equipment and method |
| CN111078912A (en) * | 2019-12-18 | 2020-04-28 | 国网上海市电力公司 | Power equipment image data warehouse and power equipment defect detection method |
| CN111078912B (en) * | 2019-12-18 | 2024-02-20 | 国网上海市电力公司 | Power equipment image data warehouse and power equipment defect detection method |
| CN111696092B (en) * | 2020-06-11 | 2023-08-25 | 深圳市华汉伟业科技有限公司 | Defect detection method and system based on feature comparison and storage medium |
| CN111696092A (en) * | 2020-06-11 | 2020-09-22 | 深圳市华汉伟业科技有限公司 | Defect detection method and system based on feature comparison and storage medium |
| CN111986178A (en) * | 2020-08-21 | 2020-11-24 | 北京百度网讯科技有限公司 | Product defect detection method and device, electronic equipment and storage medium |
| CN112026350A (en) * | 2020-09-17 | 2020-12-04 | 广州夕千科技有限公司 | Printing machine for processing electronic components |
| CN112396599A (en) * | 2020-12-04 | 2021-02-23 | 惠州高视科技有限公司 | Visual inspection method for transparent packaged IC (integrated circuit) defects |
| CN112834518A (en) * | 2021-01-06 | 2021-05-25 | 优刻得科技股份有限公司 | Particle defect detection method, system, device and medium |
| CN113111828A (en) * | 2021-04-23 | 2021-07-13 | 中国科学院宁波材料技术与工程研究所 | Three-dimensional defect detection method and system for bearing |
| 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 |
| CN114581362A (en) * | 2021-07-22 | 2022-06-03 | 正泰集团研发中心(上海)有限公司 | Photovoltaic module defect detection method and device, electronic equipment and readable storage medium |
| CN114581362B (en) * | 2021-07-22 | 2023-11-07 | 正泰集团研发中心(上海)有限公司 | Photovoltaic module defect detection method and device, electronic equipment and readable storage medium |
| CN113466261A (en) * | 2021-07-26 | 2021-10-01 | 鸿安（福建）机械有限公司 | PCB board automatic checkout device |
| CN114152627A (en) * | 2022-02-09 | 2022-03-08 | 季华实验室 | Chip circuit defect detection method and device, electronic equipment and storage medium |
| CN114708266A (en) * | 2022-06-07 | 2022-07-05 | 青岛通产智能科技股份有限公司 | Tool, method and device for detecting card defects and medium |
| TWI803393B (en) * | 2022-07-19 | 2023-05-21 | 神通資訊科技股份有限公司 | Defect detection system based on unsupervised learning and method thereof |
| CN115239734A (en) * | 2022-09-23 | 2022-10-25 | 成都数之联科技股份有限公司 | Model training method, device, storage medium, equipment and computer program product |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109871895B (en) | 2021-03-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109871895A (en) | The defect inspection method and device of circuit board | |
| CN109886950A (en) | The defect inspection method and device of circuit board | |
| CN103324937B (en) | The method and apparatus of label target | |
| CN110473173A (en) | A kind of defect inspection method based on deep learning semantic segmentation | |
| CN109886964A (en) | Circuit board defect detection method, device and equipment | |
| CN108156166A (en) | Abnormal access identification and connection control method and device | |
| EP4163889A1 (en) | System for counting quantity of game tokens | |
| CN109389599A (en) | A kind of defect inspection method and device based on deep learning | |
| CN108038107A (en) | Sentence sensibility classification method, device and its equipment based on convolutional neural networks | |
| CN109916921A (en) | Circuit board defect processing method, device and equipment | |
| CN110647875A (en) | Method for segmenting and identifying model structure of blood cells and blood cell identification method | |
| CN114577812A (en) | Method and system for training efficient quality inspection model based on ultra-small sample | |
| CN109242831A (en) | Picture quality detection method, device, computer equipment and storage medium | |
| CN109740609A (en) | A kind of gauge detection method and device | |
| CN109816048A (en) | A kind of image composition method based on attribute migration | |
| CN110084230A (en) | Vehicle body direction detection method and device based on image | |
| CN110084289A (en) | Image labeling method, device, electronic equipment and storage medium | |
| CN114862832A (en) | Method, device and equipment for optimizing defect detection model and storage medium | |
| CN105629148B (en) | The test method and test equipment of multiple modules in test circuit | |
| JP2022045688A (en) | Defect management device, method, and program | |
| CN111739037A (en) | Semantic segmentation method for indoor scene RGB-D image | |
| CN107948586A (en) | Trans-regional moving target detecting method and device based on video-splicing | |
| CN108090905B (en) | The judgment method and system of producing line exception | |
| CN104408542B (en) | Quality models and graphical defect input method and system | |
| CN109558505A (en) | Visual search method, apparatus, computer equipment and storage medium |
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 |