CN107862693A - Detection method and device for nickel foam surface defect - Google Patents
Detection method and device for nickel foam surface defect Download PDFInfo
- Publication number
- CN107862693A CN107862693A CN201711295255.5A CN201711295255A CN107862693A CN 107862693 A CN107862693 A CN 107862693A CN 201711295255 A CN201711295255 A CN 201711295255A CN 107862693 A CN107862693 A CN 107862693A
- Authority
- CN
- China
- Prior art keywords
- sub
- block
- defect
- nickel foam
- characteristic vector
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
- G01N2021/8887—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges based on image processing techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30136—Metal
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- Quality & Reliability (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
The present invention provides a kind of detection method and device for nickel foam surface defect.This method includes：Using horizontal segmentation or vertical segmentation, the pretreated surface image is divided into multiple first sub-blocks, the size of the plurality of first sub-block is N/i x N/i, and i is more than or equal to 3；The plurality of first sub-block is subjected to Surface Defect Recognition, if at least one first sub-block in the plurality of first sub-block is identified as defect sub-block, at least one first sub-block is then divided into multiple second sub-blocks using horizontal segmentation or vertical segmentation, to carry out Surface Defect Recognition to the plurality of second sub-block, wherein, the size of each second sub-block is M/j x M/j, and j is more than or equal to 3., can the defects of fast positioning foam nickel surface position provided by the present invention for the detection method of nickel foam surface defect.
Description
Technical field
The disclosed embodiment of the present invention is related to detection technique, and more specifically, is related to one kind and is used for foam nickel surface
The detection method and device of defect.
Background technology
Nickel foam is a kind of new function material by nickel metal obtain after a series of physical chemical process, is made
It is very big for the base material of Vehicular battery, influence of the nickel foam to battery performance.Foam nickel surface is in silver-gray metallic luster, form
Similar to metal sponge, required in quality surfacing, without scuffing, flawless, without it is damaged, without greasy dirt and non-oxidation.
However, in nickel foam preparation process, due to raw material, processing technology (PVD, plating, sintering etc.) etc. factor,
Foam nickel surface is caused to pollute, impression, folding line, nickel skin, crackle, line, the defects of a variety of different types such as plating leakage, these
Defect has a strong impact on the performance and quality of final finished.
Detection to nickel foam surface defect at present and identification mainly use artificial detection means, i.e., are grasped in preparation process
The nickel foam on conveyer belt is observed by the naked eye as personnel, by rule of thumb to nickel foam the defects of manually adjudicated, and then carry out
Corresponding manual handle.Which has that labor intensity is big, and efficiency is low, and subjectivity is by force and error detection rate height.Due to quality testing point
Analysis hysteresis, it is also difficult to the production process of effective on-line optimization nickel foam.
The content of the invention
, can the present invention solves the technical problem of the detection method and device provided for nickel foam surface defect
Solve the problems of the prior art, the defects of nickel foam is gone out with fast positioning, and detect the defects of its.
In order to solve the above-mentioned technical problem, one aspect of the present invention provides a kind of detection side for nickel foam surface defect
Method.This method includes：The surface image of nickel foam is obtained, and the surface image is pre-processed, wherein pretreated table
The size of face image is that N x N, N are integer；Using horizontal segmentation or vertical segmentation, the pretreated surface image is split
For multiple first sub-blocks, the size of the plurality of first sub-block is N/i x N/i, and i is more than or equal to 3；By the plurality of first sub-block
Surface Defect Recognition is carried out, if at least one first sub-block in the plurality of first sub-block is identified as defect sub-block, is utilized
At least one first sub-block is divided into multiple second sub-blocks by horizontal segmentation or vertical segmentation, to enter to the plurality of second sub-block
Row Surface Defect Recognition, wherein, the size of each second sub-block is M/j x M/j, and M is integer and M is equal to N/i, and j is more than or waited
In 3.
The beneficial effects of the invention are as follows：When getting the surface image of nickel foam, by first splitting to it, then it is right
Sub-block after segmentation is detected, the defects of fast positioning foam nickel surface position, accurate to obtain defect in nickel foam
Feature, favourable foundation is provided for subsequent treatment.
Brief description of the drawings
Fig. 1 a, Fig. 1 b, Fig. 1 c and Fig. 1 d are that nickel foam schematic surface free of surface defects, surface have crack defect respectively
Nickel foam schematic surface, surface has a nickel foam schematic surface of plating leakage defect and there is the nickel foam of pollution defect on surface
Schematic surface.
Fig. 2 is the flow chart of the detection method for nickel foam surface defect of first embodiment of the invention.
Fig. 3 a-3c are that the surface image of pretreated nickel foam is carried out to the schematic diagram after once splitting.
Fig. 4 is the flow chart of the detection method for nickel foam surface defect of second embodiment of the invention.
Fig. 5 is the flow chart of the detection method for nickel foam surface defect of third embodiment of the invention.
Fig. 6 is the flow chart of the detection method for nickel foam surface defect of fourth embodiment of the invention.
Fig. 7 is the schematic diagram of the detection means for nickel foam surface defect of first embodiment of the invention.
Fig. 8 is the schematic diagram of the detection means for nickel foam surface defect of second embodiment of the invention.
Fig. 9 a-9b are the schematic diagrames that the first sub-block is divided into multiple second sub-blocks of second embodiment of the invention.
Embodiment
Fig. 1 a, Fig. 1 b, Fig. 1 c and Fig. 1 d are that nickel foam schematic surface free of surface defects, surface have crack defect respectively
Nickel foam schematic surface, surface has a nickel foam schematic surface of plating leakage defect and there is the nickel foam of pollution defect on surface
Schematic surface.As shown in Fig. 1 a, Fig. 1 b, Fig. 1 c and Fig. 1 d, nickel foam configuration of surface is similar to metal sponge.Required in quality
Nickel foam surfacing, without scuffing, flawless, without it is damaged, without greasy dirt and non-oxidation.However, due to raw material, processing technology
(PVD, plating, sintering etc.) etc. factor, causes foam nickel surface to pollute, impression, folding line, nickel skin, crackle, line,
The defects of a variety of different types such as plating leakage, these defects have a strong impact on the performance and quality of final finished.According to lacking for nickel foam
The characteristics of falling into, of the invention proposes a kind of detection method for nickel foam surface defect.On the surface for getting nickel foam
During image, detected by first splitting to it, then to the sub-block after segmentation, lacking on fast positioning foam nickel surface
Position is fallen into, the accurate feature for obtaining defect in nickel foam, favourable foundation is provided for subsequent treatment.
Fig. 2 is the flow chart of the detection method for nickel foam surface defect of first embodiment of the invention.This method bag
Include following steps：
Step S202：The surface image of nickel foam is obtained, and the surface image is pre-processed.
The size of wherein pretreated surface image is that N x N, N are integer.In one embodiment, pretreatment includes
To down-sampling.
Step S204：Using horizontal segmentation or vertical segmentation, the pretreated surface image is divided into multiple first
Sub-block.
Wherein, the size of each first sub-block is N/i x N/i, and i is more than or equal to 3.In one embodiment, as schemed
Shown in 3a, pretreated surface image is image A.As i=3, i.e. image A is divided into 9 the first sub-blocks, and each first
The size of sub-block is N/3x N/3.In another embodiment, as shown in Figure 3 b, i=4, i.e. image A are divided into 16 first
Sub-block.In another embodiment, as shown in Figure 3 c, i=5, i.e. image A are divided into 25 the first sub-blocks.
Step S206：The plurality of first sub-block is subjected to Surface Defect Recognition, if at least one in the plurality of first sub-block
Individual first sub-block is identified as defect image (that is, the sub-block is defect sub-block), then should using horizontal segmentation or vertical segmentation
At least one first sub-block is divided into multiple second sub-blocks, to carry out Surface Defect Recognition to the plurality of second sub-block.Wherein, often
The size of individual second sub-block is M/j x M/j, and M is integer and M is equal to N/i, and j is more than or equal to 3.
Assuming that illustrated by taking the first sub-block split in Fig. 3 a as an example, to the first sub-block 1_1 to the first sub-block
3_3 is detected, wherein the result for assuming detection is：First sub-block 1_1 is the first sub-block comprising defect, i.e. defect sub-block,
Now, the first sub-block 1_1 is split respectively, respectively obtains multiple second sub-blocks.As illustrated in fig. 9, it is assumed that enter according to j=3
Row segmentation, the first sub-block 1_1 will be divided into 9 the second sub-blocks.The like, can so detect in nickel foam defect and
Its position, is advantageous to subsequent operation.
If the result of detection is the first sub-block 1_1 and the first sub-block 2_1 is the first sub-block comprising defect, i.e. defecton
Block, now, while the first sub-block 1_1 and the first sub-block 2_1 are split, multiple second sub-blocks are respectively obtained, such as Fig. 9 b institutes
Show, it is assumed that split according to j=3, the first sub-block 1_1 is divided into 9 the second sub-blocks, then this 9 second sub-blocks are entered respectively
Row detection, the like, defect and its position in nickel foam can be so detected, is advantageous to subsequent operation.
By the implementation of above-described embodiment, when getting the surface image of nickel foam, by first splitting to it, then
Sub-block after segmentation is detected, the defects of fast positioning foam nickel surface position, it is accurate to obtain lacking in nickel foam
Feature is fallen into, favourable foundation is provided for subsequent treatment.
In one embodiment, after once split, the size of the second sub-block of secondary splitting is by defect sub-block
Difference degree between non-defective sub-block determines, wherein, non-defective sub-block refers to the surface image of the nickel foam of the sub-block
In the absence of defect, i.e., normal nickel foam, and defect sub-block refers to the surface image existing defects of the nickel foam of the sub-block.By scheming
The surface defect that 1a, Fig. 1 b, Fig. 1 c and Fig. 1 d can be seen that nickel foam is generally in block distribution.In one example, if defect
Difference between sub-block and non-defective sub-block is bigger, then possible defect area account for the area of sub-block proportion it is larger, i.e. defect
The ratio of area/sub-block area is larger.Therefore, when carrying out secondary splitting to defect sub-block, j values are taken into smaller value, to cause
The size of the second sub-block split is larger., can on the contrary, if the difference between defect sub-block and non-defective sub-block is smaller
Can the defects of the area area that accounts for sub-block proportion it is smaller, i.e. the ratio of defect area/sub-block area is smaller.Therefore, to lacking
When falling into sub-block progress secondary splitting, j values are taken into higher value, to cause the size of the second split sub-block smaller.
According to another embodiment of the present invention, at least characteristic vector of each in the plurality of second sub-block is extracted；
If dissimilar between at least characteristic vector of each in the plurality of second sub-block, flaw size is more than M/j x M/
j。
As shown in figure 4, it is the detection method for nickel foam surface defect of second embodiment of the invention.In the present embodiment
In, it is on the basis of the detection method for nickel foam surface defect of first embodiment, to the sub-block split every time
Carry out surface defects detection.In the present embodiment, this method includes：
Step S402：Extract an at least characteristic vector for the plurality of first sub-block.
At least one characteristic vector can be with gray feature, edge feature, chromaticity and the textural characteristics of the first sub-block
In one or its combination.Wherein, the textural characteristics include energy, entropy, contrast and inverse difference moment.In one embodiment,
Gray level co-occurrence matrixes can be utilized to calculate textural characteristics.
In one embodiment, the mode of an at least characteristic vector for the first sub-block of extraction is：First the first sub-block is carried out
Non-downsampling Contourlet conversion, gray level co-occurrence matrixes are then asked in a different direction to the sub-block after conversion.Another
In one embodiment, extracting the mode of the characteristic vector of the first sub-block can be：First sub-block is carried out to down-sampling, to obtain phase
The half-tone information answered.In one example, the gray feature is related to gray-scale intensity function and/or gray accumulation distribution function,
And according to gray-scale intensity function and/or gray accumulation distribution function, come judge in the plurality of first sub-block it is at least one whether
For defect sub-block.That is, judged using gray-scale intensity function as a characteristic vector.According to another embodiment party of the present invention
Formula, the gray-scale intensity function of sub-block is normalized, obtains grey level probability density function, and corresponding gray accumulation
Distribution function.According to grey level probability density function and/or gray accumulation distribution function, judge in the plurality of first sub-block at least
Whether one be defect sub-block.If for example, the probability density of a certain gray level of a certain sub-block in the plurality of first sub-block with
The difference of the grey level probability density of the gray level of other sub-blocks is more than a threshold value, then judges the sub-block for defect sub-block.Or
If a certain threshold function table area of the gray accumulation distribution function of a certain sub-block in the plurality of first sub-block and other sub-blocks
The difference of the threshold function table area of the gray accumulation distribution function is more than a threshold value, then judges the sub-block for defect sub-block.
Continuation illustrates by taking Fig. 3 a as an example, in extraction the first sub-block 1_1 to the first sub-block 3_3 this 9 the first sub-blocks
During at least one characteristic vector, the characteristic vector of each first sub-block can be extracted simultaneously, can also individually extract each first
The characteristic vector of sub-block.
Step S404：First sub-block is selected from multiple first sub-blocks as sub-block to be measured.
Continuation illustrates by taking Fig. 3 a as an example, is selected from the first sub-block 1_1 to the first sub-block 3_3 in this 9 first sub-blocks
First sub-block 1_1 is as sub-block to be measured.
Step S406：By other the first sub-blocks in an at least characteristic vector for sub-block to be measured and the plurality of first sub-block
A corresponding at least characteristic vector is compared, to determine whether the sub-block to be measured is that (that is, the sub-block is defect to defect image
Sub-block), and then Surface Defect Recognition is carried out to the plurality of first sub-block.
Continuation illustrates by taking Fig. 3 a as an example, by the first sub-block 1_1 respectively with the first sub-block 1_2, the first sub-block 1_3, the
One sub-block 2_1, the first sub-block 2_2, the first sub-block 2_3, the first sub-block 3_1, the first sub-block 3_2, the first sub-block 3_3 are compared
Compared with.If these sub-blocks are normal flawless nickel foam image subblock, the characteristic vector of these image subblocks is similar, or
Difference between these characteristic vectors is in estimation range.In the present embodiment, record respectively the first sub-block 1_1 with other first
Characteristic vector difference between sub-block, if all characteristic vector differences between the first sub-block 1_1 and other first sub-blocks are located
In preset range, then judge the first sub-block 1_1 for zero defect sub-block.If between the first sub-block 1_1 and other first sub-blocks
All characteristic vector differences are not entirely in preset range, then determine whether first sub-block 1_1 is defect according to actual conditions
Sub-block.In one embodiment, if the characteristic vector difference between the first sub-block 1_1 and other first sub-blocks exceedes default model
Predetermined number outside enclosing, then judge first sub-block for defect sub-block.For example, the first sub-block 1_1 and the first sub-block 1_3 it
Between, between the first sub-block 1_1 and the first sub-block 2_3, between the first sub-block 1_1 and the first sub-block 3_3 and the first sub-block 1_1 and
Characteristic vector difference between one sub-block 3_2 exceedes preset range, then judges the first sub-block 1_1 for defect sub-block.In other realities
Apply in example, the feature that can also be had differences according to the first sub-block 1_1 characteristic vector and the characteristic vector of other the first sub-blocks
Vectorial quantity is judged.
Continuation illustrates by taking Fig. 3 a as an example, by the present embodiment, detects that the first sub-block 1_1 and the first sub-block 2_1 are
Defect sub-block, after splitting to the first sub-block 1_1 and the second sub-block 2_1, the present embodiment to segmentation after obtained by
The defects of second sub-block, is detected equally applicable, for simplicity, is not repeating herein.
As shown in figure 5, it is the detection method for nickel foam surface defect of third embodiment of the invention.In the present embodiment
In, it is on the basis of the detection method for nickel foam surface defect of first embodiment, the sub-block split is carried out
Surface defects detection.In the present embodiment, this method includes：
Step S502：Extract an at least characteristic vector for the plurality of first sub-block.
At least one characteristic vector can be with gray feature, edge feature, chromaticity and the textural characteristics of the first sub-block
In one or its combination.In one embodiment, the mode of an at least characteristic vector for the first sub-block of extraction is：First to
One sub-block carries out non-downsampling Contourlet conversion, then asks for gray scale symbiosis in a different direction to the sub-block after conversion
Matrix.In another embodiment, extracting the mode of the characteristic vector of the first sub-block can be：First sub-block is adopted downwards
Sample, to obtain corresponding half-tone information.
Continuation illustrates by taking Fig. 3 a as an example, in extraction the first sub-block 1_1 to the first sub-block 3_3 this 9 the first sub-blocks
During at least one characteristic vector, the characteristic vector of each first sub-block can be extracted simultaneously, can also individually extract each first
The characteristic vector of sub-block.
Step S504：By this, at least a characteristic vector is classified, and according to classification results, determines the plurality of first sub-block
In it is at least one whether be defect image (that is, the sub-block is defect sub-block).
, can be by an at least characteristic vector for the plurality of sub-block as sample, and to it according to an embodiment of invention
Carry out cluster analysis, prediction.For example, it is determined as in the plurality of sub-block after defect sub-block and non-defective sub-block, by these
Input sample of the characteristic vector of sub-block as the grader of extreme learning machine, probability limit learning machine Integrated Strategy is built, is built
Vertical nickel foam defect classification model., can be with after the foundation of nickel foam defect classification model according to another embodiment of invention
The characteristic vector of these sub-blocks is classified in real time, and according to classification results, determines whether these sub-blocks are defect sub-block.
In addition, continuing to illustrate by taking Fig. 3 a as an example, by the present embodiment, the first sub-block 1_1 and the first sub-block are detected
2_1 is defect sub-block, and after splitting to the first sub-block 1_1 and the second sub-block 2_1, the present embodiment is to gained after segmentation
To the second sub-block the defects of detect equally applicable, for simplicity, do not repeating herein.
As shown in fig. 6, it is the detection method for nickel foam surface defect of fourth embodiment of the invention.In the present embodiment
In, it is on the basis of the detection method for nickel foam surface defect of first embodiment, the sub-block split is carried out
Surface defects detection.In the present embodiment, this method includes：
Step S602：Extract an at least characteristic vector for the plurality of first sub-block.
At least one characteristic vector can be with gray feature, edge feature, chromaticity and the textural characteristics of the first sub-block
In one or its combination.In one embodiment, the mode of an at least characteristic vector for the first sub-block of extraction is：First to
One sub-block carries out non-downsampling Contourlet conversion, then asks for gray scale symbiosis in a different direction to the sub-block after conversion
Matrix.In another embodiment, extracting the mode of the characteristic vector of the first sub-block can be：First sub-block is adopted downwards
Sample, to obtain corresponding half-tone information.
Continuation illustrates by taking Fig. 3 a as an example, in extraction the first sub-block 1_1 to the first sub-block 3_3 this 9 the first sub-blocks
During at least one characteristic vector, the characteristic vector of each first sub-block can be extracted simultaneously, can also individually extract each first
The characteristic vector of sub-block.
Step S604：First sub-block is selected from multiple first sub-blocks as sub-block to be measured.
Continuation illustrates by taking Fig. 3 a as an example, is selected from the first sub-block 1_1 to the first sub-block 3_3 in this 9 first sub-blocks
First sub-block 1_1 is as sub-block to be measured.
Step S606：By an at least characteristic vector for sub-block to be measured compared with an at least standard feature vector, with right
The plurality of first sub-block carries out Surface Defect Recognition.
The component in each standard feature vector in wherein at least one standard feature vector is by the plurality of first sub-block
In other the first sub-blocks corresponding characteristic vector respective components weighted average and obtain.Continuation illustrates by taking Fig. 3 a as an example,
Assuming that getting each first sub-block has a characteristic vector (that is, characteristic vector x (i, j)), each of which characteristic vector
Including two components (that is, i and j), then the two points of other the first sub-blocks in addition to the first sub-block 1_1 are calculated respectively
The weighted average of amount.Then, by the i in the first sub-block 1_1 characteristic vector and the weighted average of other 8 the first sub-blocks
It is compared, if difference between the two is within preset range, judges the first sub-block 1_1 for zero defect sub-block, otherwise,
Judge the first sub-block 1_1 for defect sub-block.In one embodiment, if some characteristic vector is the multi-parameter that matrix represents,
The matrix that corresponding average value refers to the average value of each parameter and formed represents.
Continuation illustrates by taking Fig. 3 a as an example, by the present embodiment, detects that the first sub-block 1_1 and the first sub-block 2_1 are
Defect sub-block, after splitting to the first sub-block 1_1 and the second sub-block 2_1, the present embodiment to segmentation after obtained by
The defects of second sub-block, is detected equally applicable, for simplicity, is not repeating herein.
As shown in fig. 7, it is the detection means for nickel foam surface defect of first embodiment of the invention.The device includes
One or more electronic circuits 710.
Electronic circuit 710 controls the operation of the detection means, and it can be realized that processor can also be referred to as by processor
CPU (Central Processing Unit, CPU).Electronic circuit 710 is probably a kind of IC chip,
Disposal ability with signal.Electronic circuit 710 can also be general processor, digital signal processor (DSP), special integrated
Circuit (ASIC), ready-made programmable gate array (FPGA) either other PLDs, discrete gate or transistor logic
Device, discrete hardware components.General processor can be microprocessor or the processor can also be any conventional processing
Device etc..
Electronic circuit 710 be used for computer instructions with realize first embodiment of the invention into fourth embodiment it is any
The detection method that individual and any combination not conflicted is provided.
As shown in figure 8, it is the detection means for nickel foam surface defect of second embodiment of the invention.The device includes
Memory 810, memory 810 are stored with computer instruction, and the computer instruction realizes first embodiment of the invention when being performed
The detection method provided to any of fourth embodiment and any combination not conflicted.
Memory 810 can include read-only storage (ROM, Read-Only Memory), random access memory (RAM,
Random Access Memory), flash memory (Flash Memory), hard disk, CD etc..
Those skilled in the art is apparent from, and device and method can be made while teachings of the present invention content is kept
Many modification and variation.Therefore, disclosure above should be considered as only being limited by the scope of following claims.
Claims (10)
1. a kind of detection method for nickel foam surface defect, it is characterised in that this method includes：
The surface image of nickel foam is obtained, and the surface image is pre-processed, wherein the chi of pretreated surface image
Very little is that N x N, N are integer；
Using horizontal segmentation or vertical segmentation, the pretreated surface image is divided into multiple first sub-blocks, the plurality of
The size of one sub-block is N/i x N/i, and i is more than or equal to 3；
The plurality of first sub-block is subjected to Surface Defect Recognition, if at least one first sub-block in the plurality of first sub-block is known
Not Wei defect sub-block, then at least one first sub-block is divided into multiple second sub-blocks using horizontal segmentation or vertical segmentation,
To carry out Surface Defect Recognition to the plurality of second sub-block, wherein, the size of each second sub-block is that M/j x M/j, M are integer
And M is equal to N/i, j is more than or equal to 3.
2. the detection method according to claim 1 for nickel foam surface defect, it is characterised in that
Extract an at least characteristic vector for the plurality of first sub-block；
First sub-block is selected from the plurality of first sub-block as sub-block to be measured；
An at least characteristic vector for sub-block to be measured is corresponding with other the first sub-blocks in the plurality of first sub-block at least
One characteristic vector is compared, and to determine whether the sub-block to be measured is defect sub-block, and then carries out table to the plurality of first sub-block
Planar defect identifies.
3. the detection method according to claim 1 for nickel foam surface defect, it is characterised in that
Extract an at least characteristic vector for the plurality of first sub-block；
By this, at least a characteristic vector is classified, and according to classification results, is determined at least one in the plurality of first sub-block
Whether it is defect sub-block.
4. the detection method according to claim 1 for nickel foam surface defect, it is characterised in that
Extract an at least characteristic vector for the plurality of first sub-block；
First sub-block is selected from multiple first sub-blocks as sub-block to be measured；
By an at least characteristic vector for sub-block to be measured compared with an at least standard feature vector, with to the plurality of first sub-block
Surface Defect Recognition is carried out, wherein the component in each standard feature vector is by other the first sub-blocks in the plurality of first sub-block
The respective components weighted average of corresponding characteristic vector and obtain.
5. the detection method for nickel foam surface defect according to claims 2 to 4, it is characterised in that
At least a characteristic vector is related to one or more of gray feature, edge feature, chromaticity, textural characteristics for this.
6. the detection method according to claim 5 for nickel foam surface defect, it is characterised in that the textural characteristics bag
Energy, entropy, contrast and inverse difference moment are included, and the textural characteristics are calculated using gray level co-occurrence matrixes.
7. the detection method according to claim 5 for nickel foam surface defect, it is characterised in that the gray feature with
Gray-scale intensity function and/or gray accumulation distribution function are related, and according to gray-scale intensity function/or gray accumulation distribution function,
Come judge in the plurality of first sub-block it is at least one whether be defect sub-block.
8. the detection method according to claim 1 for nickel foam surface defect, it is characterised in that according to defect sub-block
With the difference degree between non-defective sub-block, j value is determined.
9. the image partition method according to claim 1 for nickel foam surface defects detection, it is characterised in that
Extract an at least characteristic vector for each in the plurality of second sub-block；If in the plurality of second sub-block each
Dissimilar between an at least characteristic vector, then flaw size is more than M/j x M/j.
10. a kind of detection means for nickel foam surface defect, it is characterised in that the device includes one or more electronics electricity
Road, for performing storage program in memory to realize the method described in claim 1-8 any one of kind.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711295255.5A CN107862693B (en) | 2017-12-08 | 2017-12-08 | Method and device for detecting surface defects of foamed nickel |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711295255.5A CN107862693B (en) | 2017-12-08 | 2017-12-08 | Method and device for detecting surface defects of foamed nickel |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107862693A true CN107862693A (en) | 2018-03-30 |
| CN107862693B CN107862693B (en) | 2021-10-08 |
Family
ID=61705259
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711295255.5A Active CN107862693B (en) | 2017-12-08 | 2017-12-08 | Method and device for detecting surface defects of foamed nickel |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107862693B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108564624A (en) * | 2018-04-23 | 2018-09-21 | 湘潭大学 | A kind of nickel foam image pre-processing method based on reversed P-M diffusions |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103822932A (en) * | 2014-02-27 | 2014-05-28 | 清华大学 | X-ray real-time image weld defect detection method based on multi-scale filtering operator |
| CN103886602A (en) * | 2014-03-28 | 2014-06-25 | 重庆大学 | Radial image deflect detecting method based on veins |
| CN105069459A (en) * | 2015-08-18 | 2015-11-18 | 电子科技大学 | Surface feature type extracting method for high-resolution SAR image |
| CN106373124A (en) * | 2016-09-21 | 2017-02-01 | 哈尔滨工业大学 | Gray-level co-occurrence matrix and RANSAC-based industrial product surface defect visual detection method |
| CN106600593A (en) * | 2016-12-19 | 2017-04-26 | 福州大学 | Aluminum ceramic ball surface detect detection method |
| CN106651856A (en) * | 2016-12-31 | 2017-05-10 | 湖南文理学院 | Detection method for foamed nickel surface defects |
| US20170186144A1 (en) * | 2015-12-28 | 2017-06-29 | Powerchip Technology Corporation | Method, image processing system and computer-readable recording medium for item defect inspection |
| CN106981069A (en) * | 2017-03-31 | 2017-07-25 | 卢涵宇 | A kind of high resolution image disaster object fast partition method |
-
2017
- 2017-12-08 CN CN201711295255.5A patent/CN107862693B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103822932A (en) * | 2014-02-27 | 2014-05-28 | 清华大学 | X-ray real-time image weld defect detection method based on multi-scale filtering operator |
| CN103886602A (en) * | 2014-03-28 | 2014-06-25 | 重庆大学 | Radial image deflect detecting method based on veins |
| CN105069459A (en) * | 2015-08-18 | 2015-11-18 | 电子科技大学 | Surface feature type extracting method for high-resolution SAR image |
| US20170186144A1 (en) * | 2015-12-28 | 2017-06-29 | Powerchip Technology Corporation | Method, image processing system and computer-readable recording medium for item defect inspection |
| CN106373124A (en) * | 2016-09-21 | 2017-02-01 | 哈尔滨工业大学 | Gray-level co-occurrence matrix and RANSAC-based industrial product surface defect visual detection method |
| CN106600593A (en) * | 2016-12-19 | 2017-04-26 | 福州大学 | Aluminum ceramic ball surface detect detection method |
| CN106651856A (en) * | 2016-12-31 | 2017-05-10 | 湖南文理学院 | Detection method for foamed nickel surface defects |
| CN106981069A (en) * | 2017-03-31 | 2017-07-25 | 卢涵宇 | A kind of high resolution image disaster object fast partition method |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108564624A (en) * | 2018-04-23 | 2018-09-21 | 湘潭大学 | A kind of nickel foam image pre-processing method based on reversed P-M diffusions |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107862693B (en) | 2021-10-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101510262B (en) | Automatic measurement method for separated-out particles in steel and morphology classification method thereof | |
| Prill et al. | Morphological segmentation of FIB‐SEM data of highly porous media | |
| CN108022233A (en) | A kind of edge of work extracting method based on modified Canny operators | |
| CN104794491A (en) | Fuzzy clustering steel plate surface defect detection method based on pre classification | |
| CN107085846A (en) | Surface Flaw image-recognizing method | |
| CN109215009B (en) | Continuous casting billet surface image defect detection method based on deep convolution neural network | |
| CN102494976A (en) | Method for automatic measurement and morphological classification statistic of ultra-fine grain steel grains | |
| CN106683076A (en) | Texture feature clustering-based locomotive wheelset tread damage detection method | |
| CN104880389A (en) | Mixed crystal degree automatic measurement and fine classification method for steel crystal grains, and system thereof | |
| CN111724335A (en) | Detection method and detection system | |
| CN102663422B (en) | Floor layer classification method based on color characteristic | |
| CN111028229A (en) | Metal or coating corrosion detection method based on image processing technology | |
| CN105303200B (en) | Face identification method for handheld device | |
| US11913912B2 (en) | Hole expansion ratio testing device, hole expansion ratio testing method, and operating program | |
| CN110910444A (en) | RES scale shale equivalent three-dimensional pore parameter rapid extraction method | |
| CN108008006A (en) | A kind of weld defect detection method, device, equipment and system | |
| CN114743102A (en) | Furniture board oriented flaw detection method, system and device | |
| CN103886332A (en) | Method for detecting and identifying defects of metallic meshes | |
| CN112805719A (en) | Classifying defects in semiconductor samples | |
| CN107862693A (en) | Detection method and device for nickel foam surface defect | |
| Sun et al. | Post-slicing inspection of silicon wafers using the HJ-PSO algorithm under machine vision | |
| CN114972330A (en) | Workpiece surface roughness detection optimization method based on improved histogram homogenization algorithm | |
| CN114359276A (en) | Steel die blanking optimization scheme obtaining method based on pockmark defects | |
| CN108122227A (en) | For the detection method of nickel foam surface defect | |
| JP3652589B2 (en) | Defect inspection equipment |
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 | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20180330 Assignee: CHANGDE LYRUN MATERIAL Co.,Ltd. Assignor: HUNAN University OF ARTS AND SCIENCE Contract record no.: X2023980036342 Denomination of invention: Detection method and device for surface defects of foam nickel Granted publication date: 20211008 License type: Common License Record date: 20230608 |
|
| EE01 | Entry into force of recordation of patent licensing contract |