CN107730500A - A kind of ceramic tile texture detection, system, device and readable storage medium storing program for executing - Google Patents
A kind of ceramic tile texture detection, system, device and readable storage medium storing program for executing Download PDFInfo
- Publication number
- CN107730500A CN107730500A CN201711058538.8A CN201711058538A CN107730500A CN 107730500 A CN107730500 A CN 107730500A CN 201711058538 A CN201711058538 A CN 201711058538A CN 107730500 A CN107730500 A CN 107730500A
- Authority
- CN
- China
- Prior art keywords
- training
- image
- training sample
- ceramic tile
- network model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/40—Analysis of texture
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30132—Masonry; Concrete
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Image Processing (AREA)
- Chemical & Material Sciences (AREA)
- Evolutionary Biology (AREA)
- Quality & Reliability (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Image Analysis (AREA)
- Bioinformatics & Cheminformatics (AREA)
Abstract
This application discloses a kind of ceramic tile texture detection, including：Obtain the first training sample and the second training sample；Wherein, the first training sample includes training image and corresponding training characteristics value, and the second training sample includes training characteristics value；Using the first training sample, depth convolutional network model, depth convolutional network model after being trained are trained；Using the second training sample, SVM classifier is built；Obtain testing image；By depth convolutional network model after testing image input training, corresponding characteristic value to be measured is obtained；Classification judgement is carried out to characteristic value to be measured by SVM classifier.The present invention solve thes problems, such as that original template identification time-consuming long, standard mechanical, parameter are difficult to adjustment, quickly ceramic tile surface texture can well be judged, reduce product defect rate.The application further correspondingly discloses a kind of ceramic tile skin texture detection system, device and readable storage medium storing program for executing.
Description
Technical field
The present invention relates to ceramic tile field, more particularly to a kind of ceramic tile texture detection, system, device and readable storage medium
Matter.
Background technology
Now in society, ceramic tile is the most important material of construction and decoration, and expensive, high-grade building construction often uses
Advanced ceramic tile.However, 21 century, ceramic industry generally subjects the pressure of cost increase, environmental protection, anti-dumping etc.,
The particularly homogeneity of industry product, exacerbate the competition of industry various brands.In this case, ceramic tile enterprise increasingly focuses on
Product is made, and more focuses on intelligentized production, to attract more businessmans to be traded.
Increasing intelligence system is applied to ceramic tile industry, such as automatic paving line technology, glaze spraying system of robot now
System etc., more preferable enterprise be even more from Raw material processing to glaze line equipment all using digital intelligent, so, the production of ceramic tile industry
It is intelligent progressively to turn into production main flow, no longer it is the fixed production scale of the past.
A most indispensable part is exactly detection in intelligent production process, the fine or not very big journey of detection
The level for embodying an Intelligent Production System of degree.Conventional quality testing mode is often artificial treatment mode, old employee
Detectability it is more many by force than the detectability of new employee, very big human error in detection process be present.Part is looked forward to now
Industry gradually makes the transition after intelligent production, and detection is also transformed into intelligentized detection, including the deformation of detection ceramic tile, size are inclined
Difference, cracking, aberration, ooze that floral pattern is fuzzy, surface texture etc..
Present detection of the ceramic tile factories for ceramic tile surface texture is also in artificial detection process, by experienced employee
Skin texture detection is carried out to ceramic tile, because whether texture there is defect not possess specific judgment criteria, can only be relied on experienced
Employee goes to detect one by one.By the resolution of worker's naked eyes, this method is not only poorly efficient but also lower and noisy working long hours
Visual fatigue occurs in worker in environment, it may appear that flase drop or missing inspection, will can largely improve the substandard products of enterprise's production
Rate.
Now, texture is also identified by way of template matches based on machine vision, two is entered with money ceramic tile
Row carries out image difference processing to judge texture quality after taking pictures.But due to the diversity of ceramic tile texture, pass through template matches
The accuracy rate that method is detected is not high, moreover, having particular requirement to taking the photograph phase device and picture position, operates relatively complicated.On
The feature that traditional image-recognizing method only extracts the part of representative of image, such as SIFT and SURF are stated, is had certain
Limitation, some processes still need artificial selection.
The content of the invention
In view of this, it is an object of the invention to provide ceramic tile texture detection, system, the device of a kind of efficiently and accurately
And readable storage medium storing program for executing.Its concrete scheme is as follows：
A kind of ceramic tile texture detection, including：
Obtain the first training sample and the second training sample；Wherein, first training sample includes training image and right
The training characteristics value answered, second training sample include training characteristics value；
Using first training sample, depth convolutional network model, depth convolutional network model after being trained are trained；
Using second training sample, SVM classifier is built；
Obtain testing image；
By depth convolutional network model after the testing image input training, corresponding characteristic value to be measured is obtained；
Classification judgement is carried out to the characteristic value to be measured by the SVM classifier.
Preferably, the process for obtaining the first training sample includes：
Obtain first training sample；Wherein, first training sample includes training image and corresponding training is special
Value indicative；
Image preprocessing is carried out to the training image；
Accordingly, the process for obtaining test image includes：
Obtain testing image；
Described image pretreatment is carried out to the testing image.
Preferably, described image pretreatment includes：
Adjust image size, and/or image gray processing, and/or image denoising sound.
Preferably, it is described to utilize first training sample, depth convolutional network model is trained, depth is rolled up after being trained
The process of product network model, including：
A:The training characteristics value is normalized, determines learning parameter；
B:Using first training sample, implicit layer matrix and the output layer square of the depth convolutional network model is calculated
Battle array；
C:Calculate hidden layer modified weight amount and output layer modified weight amount；
D:According to the hidden layer modified weight amount and the output layer modified weight amount, update the implicit layer matrix and
The output layer matrix；
E:Judge whether the implicit layer matrix and the output layer matrix meet the learning parameter, if it is,
Depth convolutional network model after to the training, if it is not, then return to step B.
Preferably, the learning parameter includes：
Learning efficiency, and/or precision, and/or study number.
Accordingly, the invention also discloses a kind of ceramic tile skin texture detection system, including：
First acquisition module, for obtaining the first training sample and the second training sample；Wherein, first training sample
Including training image and corresponding training characteristics value, second training sample includes training characteristics value；
Model learning module, for utilizing first training sample, depth convolutional network model is trained, after obtaining training
Depth convolutional network model；
Classification learning module, for utilizing second training sample, build SVM classifier；
Second acquisition module, for obtaining testing image；
Test module, for the testing image to be inputted into depth convolutional network model after the training, obtain corresponding to
Characteristic value to be measured；
Determination module, for carrying out classification judgement to the characteristic value to be measured by the SVM classifier.
Preferably, the ceramic tile skin texture detection system also includes：
Pretreatment module, for carrying out image preprocessing to the training image and the testing image.
Preferably, described image pretreatment includes：
Adjust image size, and/or image gray processing, and/or image denoising sound.
Accordingly, the invention also discloses a kind of ceramic tile skin texture detection device, including：
Memory, for storing computer program；
Processor, the ceramic tile texture detection as any one of above is realized during for performing the computer program
The step of.
Accordingly, the invention also discloses a kind of readable storage medium storing program for executing, computer is stored with the readable storage medium storing program for executing
Program, the step of the ceramic tile texture detection as any one of above is realized when the computer program is executed by processor
Suddenly.
The invention discloses a kind of ceramic tile texture detection, including：Obtain the first training sample and the second training sample；
Wherein, first training sample includes training image and corresponding training characteristics value, and second training sample includes training
Characteristic value；Using first training sample, depth convolutional network model, depth convolutional network model after being trained are trained；
Using second training sample, SVM classifier is built；Obtain testing image；After the testing image is inputted into the training
Depth convolutional network model, obtain corresponding characteristic value to be measured；The characteristic value to be measured is divided by the SVM classifier
Class judges.Deep learning is combined by the present invention with ceramic tile surface texture recognition, gives full play to the excellent of deep learning self-teaching
Gesture, solve the problems, such as that original template identification time-consuming long, standard mechanical, parameter are difficult to adjustment, can be quickly to ceramic tile table
Face texture is well judged, reduces product defect rate.
Brief description of the drawings
In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing
There is the required accompanying drawing used in technology description to be briefly described, it should be apparent that, drawings in the following description are only this
The embodiment of invention, for those of ordinary skill in the art, on the premise of not paying creative work, can also basis
The accompanying drawing of offer obtains other accompanying drawings.
Fig. 1 is a kind of step flow chart of ceramic tile texture detection in the embodiment of the present invention；
Fig. 2 is a kind of structure distribution figure of ceramic tile skin texture detection device in the embodiment of the present invention.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Site preparation describes, it is clear that described embodiment is only part of the embodiment of the present invention, rather than whole embodiments.It is based on
Embodiment in the present invention, those of ordinary skill in the art are obtained every other under the premise of creative work is not made
Embodiment, belong to the scope of protection of the invention.
It is shown in Figure 1 the embodiment of the invention discloses a kind of ceramic tile texture detection, including：
S1：Obtain the first training sample and the second training sample；
Wherein, first training sample includes training image and corresponding training characteristics value, second training sample
Including training characteristics value；
It is understood that the training characteristics value in the first training sample can with the training characteristics value in the second training sample
, can also be different to be identical.
S2：Using first training sample, depth convolutional network model, depth convolutional network mould after being trained are trained
Type；
It is understood that depth convolutional network model can be obtained corresponding to image by the image of input after training
Characteristic value.The process of specific training depth convolutional network model is shown in following steps：
A:The training characteristics value is normalized, determines learning parameter；
B:Using first training sample, implicit layer matrix and the output layer square of the depth convolutional network model is calculated
Battle array；
C:Calculate hidden layer modified weight amount and output layer modified weight amount；
D:According to the hidden layer modified weight amount and the output layer modified weight amount, update the implicit layer matrix and
The output layer matrix；
E:Judge whether the implicit layer matrix and the output layer matrix meet the learning parameter, if it is,
Depth convolutional network model after to the training, if it is not, then return to step B.
Wherein, above-mentioned learning parameter can include learning efficiency, and/or precision, and/or study number.Certainly can be with
Set other be used for train parameter.
In fact, the first training sample includes two classes, one kind is the image of ceramic tile surface texture preferably, and one kind is ceramic tile table
Face texture is bad image.The quality of ceramic tile texture first is judged to obtain by the experienced worker's auxiliary of factory.Build depth
Convolutional network, including input layer, hidden layer and output layer, comprise the following steps：(a) input layer input original image；(b) meter is passed through
Calculate the error between the output of the depth convolutional network and concrete class label, adjusted by error backpropagation algorithm described in
The weights and bias term of each layer of depth convolutional network, until network stabilization or reach set maximum iteration.(c) export
Layer exports representative combinations of features, includes the edge feature combination of image, the basic configuration combinations of features of image, image
Color character combination, for the later stage classify.
It is understood that convolutional network is inherently a kind of mapping for being input to output, it can learn largely
Mapping relations between input and output, without the accurate mathematic(al) representation between any input and output, it is only necessary to
Convolutional network is trained with known pattern, network just has the mapping ability between input and output.Its training sample is
By shaped like：The vector of input vector-preferable output vector is by training image-corresponding instruction in of the invention to composition
Practice the vector of characteristic value to forming.All these vectors are right, should all be derived from the reality " fortune for the system that network will simulate
OK " result.They can gather to come from actual motion system.Before training is started, all power should all be used
Different small random numbers are initialized." small random number " is used for ensureing that network will not enter saturation state because weights are excessive,
So as to cause failure to train；" difference " is used for ensureing that network can normally learn.If in fact, gone initially with identical number
Change weight matrix, then network impotentia learns.
S3：Using second training sample, structure SVM (Support Vector Machine, SVMs) divides
Class device；
It is understood that SVM classifier (SVMs) is based on the general-purpose machinery under Statistical Learning Theory framework
Learning method, initially put forward for two class classification problems, it has the advantages of simple in construction, generalization ability is strong.
In the present invention, it is trained in training process using the ceramic tile training sample chosen by hand, obtains a svm classifier
Device, then utilized in identification process to various features, including：Whether ceramic tile surface aberration, ceramic tile surface texture are consistent, whether there is bright
The ROI (Region of Interest, area-of-interest) that the segmentation such as aobvious dislocation or broken string phenomenon obtains, is obtained using training
Grader carries out Classification and Identification.
Wherein, using tile image aberration feature, the tool being trained to corresponding ceramic tile surface Chromatism classification device
Body step can be as follows：
Aberration judgement is carried out to training sample image, judged according to the gray value of pixel in entire image.Will instruction
Practice sample image and be first converted into gray-scale map, then count gray value given threshold value d1's (such as being set to 60≤d1≤100)
The number of pixel；If meeting, the pixel number of condition accounts for the percentage of whole image pixel number more than given threshold value
D2 (such as being set to d2=0.9), then judge that training sample image is aberration small sample；Otherwise aberration large sample is determined that it is；Most
Afterwards SVM classifier 1, the set training svm classifier that the big sample of aberration is formed are trained with the set of the small sample composition of aberration
Device 2.
Wherein, using the shape facility of ceramic tile texture image, corresponding ceramic tile surface textured pattern grader is instructed
White silk can be as follows：
The feature of its textured pattern is extracted to training sample image, first calculates the textural characteristics of a certain high-quality tile image
Minimum bounding box area and girth, it is summarized in standard feature vector；If the characteristic vector difference with remaining ceramic tile texture image
More than given threshold value d3, then judge that training sample image is deformation texture sample；Otherwise it is determined as texture without deformation sample；With
SVM classifier 3, SVM points of the set training that sample of the texture without deformation is formed are trained in the set that the sample of deformation texture is formed
Class device 4.
In Classification and Identification afterwards, corresponding classification is assigned to after aberration and surface texture feature judgement are carried out to ROI
Device.If a ROI meet multiple characteristic conditions (as be both aberration it is big and deformation texture), it is assigned to respectively
Then acquired results are judged by corresponding SVM2 and SVM3, decision method includes being judged according to confidence level or method of voting,
Final output recognition result.
It is understood that due to not associated between step S2, S3, therefore the action for this two step is successively, can be with
Do not require.
S4：Obtain testing image；
Further, for the testing image in the training image in the first training sample and step S4, after acquisition
Image preprocessing can also be carried out, can specifically include adjustment image size, and/or image gray processing, and/or image denoising
Sound, main purpose are for exclusive PCR amount, obtain apparent clear and definite input quantity.
S5：By depth convolutional network model after the testing image input training, corresponding characteristic value to be measured is obtained；
As long as it is understood that preceding four step is completed before step S5, step S4 not with step S1, S2,
S3 priority association.
S6：Classification judgement is carried out to the characteristic value to be measured by the SVM classifier.
It is understood that training after depth convolutional network model input parameter when testing image, export feature to be measured
Value, characteristic value to be measured during the input parameter of SVM classifier, output category result.Wherein, classification results represent ceramic tile surface texture
Quality, typically go to judge by ceramic tile surface aberration, whether clear this 2 points of pattern texture in practical operation.
The invention discloses a kind of ceramic tile texture detection, including：Obtain the first training sample and the second training sample；
Wherein, first training sample includes training image and corresponding training characteristics value, and second training sample includes training
Characteristic value；Using first training sample, depth convolutional network model, depth convolutional network model after being trained are trained；
Using second training sample, SVM classifier is built；Obtain testing image；After the testing image is inputted into the training
Depth convolutional network model, obtain corresponding characteristic value to be measured；The characteristic value to be measured is divided by the SVM classifier
Class judges.Deep learning is combined by the present invention with ceramic tile surface texture recognition, gives full play to the excellent of deep learning self-teaching
Gesture, solve the problems, such as that original template identification time-consuming long, standard mechanical, parameter are difficult to adjustment, can be quickly to ceramic tile table
Face texture is well judged, reduces product defect rate.
Accordingly, it is shown in Figure 2 the embodiment of the invention also discloses a kind of ceramic tile skin texture detection system, including：
First acquisition module 01, for obtaining the first training sample and the second training sample；Wherein, the first training sample
This includes training image and corresponding training characteristics value, and second training sample includes training characteristics value；
Model learning module 02, for utilizing first training sample, depth convolutional network model is trained, is trained
Depth convolutional network model afterwards；
Classification learning module 03, for utilizing second training sample, build SVM classifier；
Second acquisition module 04, for obtaining testing image；
Test module 05, for the testing image to be inputted into depth convolutional network model after the training, obtain correspondingly
Characteristic value to be measured；
Determination module 06, for carrying out classification judgement to the characteristic value to be measured by the SVM classifier.
Further, the ceramic tile skin texture detection system can also include：
Pretreatment module, for carrying out image preprocessing to the training image and the testing image.
Specifically, described image pretreatment can include：
Adjust image size, and/or image gray processing, and/or image denoising sound.
Accordingly, the embodiment of the invention also discloses a kind of ceramic tile skin texture detection device, including：
Memory, for storing computer program；
Processor, the ceramic tile texture detection as described in foregoing embodiments is realized during for performing the computer program
The step of.
Wherein, the detail about the ceramic tile skin texture detection device may be referred to ceramic tile texture in above-described embodiment and examine
Survey method, here is omitted.
Accordingly, the invention also discloses a kind of readable storage medium storing program for executing, computer is stored with the readable storage medium storing program for executing
Program, the step of the ceramic tile texture detection as any one of above is realized when the computer program is executed by processor
Suddenly.
Wherein, the detail about the readable storage medium storing program for executing may be referred to ceramic tile skin texture detection side in above-described embodiment
Method, here is omitted.
Finally, it is to be noted that, herein, such as first and second or the like relational terms be used merely to by
One entity or operation make a distinction with another entity or operation, and not necessarily require or imply these entities or operation
Between any this actual relation or order be present.Moreover, term " comprising ", "comprising" or its any other variant meaning
Covering including for nonexcludability, so that process, method, article or equipment including a series of elements not only include that
A little key elements, but also the other element including being not expressly set out, or also include for this process, method, article or
The intrinsic key element of equipment.In the absence of more restrictions, the key element limited by sentence "including a ...", is not arranged
Except other identical element in the process including the key element, method, article or equipment being also present.
A kind of ceramic tile texture detection provided by the present invention, system, device and readable storage medium storing program for executing are carried out above
It is discussed in detail, specific case used herein is set forth to the principle and embodiment of the present invention, above example
Explanation be only intended to help understand the present invention method and its core concept；Meanwhile for those of ordinary skill in the art,
According to the thought of the present invention, there will be changes in specific embodiments and applications, in summary, in this specification
Appearance should not be construed as limiting the invention.
Claims (10)
- A kind of 1. ceramic tile texture detection, it is characterised in that including：Obtain the first training sample and the second training sample；Wherein, first training sample includes training image and corresponding Training characteristics value, second training sample include training characteristics value；Using first training sample, depth convolutional network model, depth convolutional network model after being trained are trained；Using second training sample, SVM classifier is built；Obtain testing image；By depth convolutional network model after the testing image input training, corresponding characteristic value to be measured is obtained；Classification judgement is carried out to the characteristic value to be measured by the SVM classifier.
- 2. ceramic tile texture detection according to claim 1, it is characterised in thatThe process for obtaining the first training sample includes：Obtain first training sample；Wherein, first training sample includes training image and corresponding training characteristics value；Image preprocessing is carried out to the training image；The process for obtaining test image includes：Obtain testing image；Described image pretreatment is carried out to the testing image.
- 3. ceramic tile texture detection according to claim 2, it is characterised in that described image pretreatment includes：Adjust image size, and/or image gray processing, and/or image denoising sound.
- 4. according to any one of claims 1 to 3 ceramic tile texture detection, it is characterised in that described to utilize described first Training sample, training depth convolutional network model, the process of depth convolutional network model after being trained, including：A:The training characteristics value is normalized, determines learning parameter；B:Using first training sample, calculate the implicit layer matrix of the depth convolutional network model and export layer matrix；C:Calculate hidden layer modified weight amount and output layer modified weight amount；D:According to the hidden layer modified weight amount and the output layer modified weight amount, the implicit layer matrix and described is updated Export layer matrix；E:Judge whether the implicit layer matrix and the output layer matrix meet the learning parameter, if it is, obtaining institute Depth convolutional network model after training is stated, if it is not, then return to step B.
- 5. ceramic tile texture detection according to claim 4, it is characterised in that the learning parameter includes：Learning efficiency, and/or precision, and/or study number.
- A kind of 6. ceramic tile skin texture detection system, it is characterised in that including：First acquisition module, for obtaining the first training sample and the second training sample；Wherein, first training sample includes Training image and corresponding training characteristics value, second training sample include training characteristics value；Model learning module, for utilizing first training sample, train depth convolutional network model, depth after being trained Convolutional network model；Classification learning module, for utilizing second training sample, build SVM classifier；Second acquisition module, for obtaining testing image；Test module, for the testing image to be inputted into depth convolutional network model after the training, obtain corresponding to it is to be measured Characteristic value；Determination module, for carrying out classification judgement to the characteristic value to be measured by the SVM classifier.
- 7. ceramic tile skin texture detection system according to claim 6, it is characterised in that also include：Pretreatment module, for carrying out image preprocessing to the training image and the testing image.
- 8. ceramic tile skin texture detection system according to claim 7, it is characterised in that described image pretreatment includes：Adjust image size, and/or image gray processing, and/or image denoising sound.
- A kind of 9. ceramic tile skin texture detection device, it is characterised in that including：Memory, for storing computer program；Processor, the ceramic tile skin texture detection side as described in any one of claim 1 to 5 is realized during for performing the computer program The step of method.
- 10. a kind of readable storage medium storing program for executing, it is characterised in that computer program, the meter are stored with the readable storage medium storing program for executing Realized when calculation machine program is executed by processor as described in any one of claim 1 to 5 the step of ceramic tile texture detection.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711058538.8A CN107730500A (en) | 2017-11-01 | 2017-11-01 | A kind of ceramic tile texture detection, system, device and readable storage medium storing program for executing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711058538.8A CN107730500A (en) | 2017-11-01 | 2017-11-01 | A kind of ceramic tile texture detection, system, device and readable storage medium storing program for executing |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107730500A true CN107730500A (en) | 2018-02-23 |
Family
ID=61221463
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711058538.8A Pending CN107730500A (en) | 2017-11-01 | 2017-11-01 | A kind of ceramic tile texture detection, system, device and readable storage medium storing program for executing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107730500A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109051106A (en) * | 2018-06-28 | 2018-12-21 | 广东科达洁能股份有限公司 | A kind of ceramic tile potted flower color packing method and system |
| CN109523505A (en) * | 2018-09-18 | 2019-03-26 | 深圳市智信精密仪器有限公司 | A kind of ceramic brick surface patterned defect detection method based on machine vision |
| CN109697303A (en) * | 2018-11-16 | 2019-04-30 | 台州旭日环境科技有限公司 | A kind of building brick sticking scheme generation method towards plane special-shaped region |
| CN109829496A (en) * | 2019-01-30 | 2019-05-31 | 广州市载道信息科技有限公司 | A kind of physical measurement classification method and equipment |
| CN110163244A (en) * | 2019-04-04 | 2019-08-23 | 睿视智觉（厦门）科技有限公司 | A kind of ceramic tile texture classifying method and device |
| CN110702839A (en) * | 2019-09-17 | 2020-01-17 | 珠海格力电器股份有限公司 | Method and device for detecting gas pollutants |
| CN110930352A (en) * | 2019-02-25 | 2020-03-27 | 研祥智能科技股份有限公司 | Object color difference defect detection method and device, computer equipment and storage medium |
| CN112633393A (en) * | 2020-12-29 | 2021-04-09 | 北京理工大学重庆创新中心 | Automatic classification method and device for ceramic tile textures |
| CN113674305A (en) * | 2021-08-25 | 2021-11-19 | 茶陵县强强陶瓷有限公司 | Intelligent blank storage system applied to ceramic production and control method |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN204116238U (en) * | 2014-06-24 | 2015-01-21 | 中南大学 | A kind of ceramic tile texture on-line checkingi sorter |
| CN106599848A (en) * | 2016-12-16 | 2017-04-26 | 南京理工大学 | Depth visual feature and support vector machine-based terrain texture recognition algorithm |
| US20170116728A1 (en) * | 2015-10-22 | 2017-04-27 | International Business Machines Corporation | Viewpoint recognition in computer tomography images |
| CN107192716A (en) * | 2017-04-26 | 2017-09-22 | 广东工业大学 | A kind of workpiece, defect quick determination method based on contour feature |
-
2017
- 2017-11-01 CN CN201711058538.8A patent/CN107730500A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN204116238U (en) * | 2014-06-24 | 2015-01-21 | 中南大学 | A kind of ceramic tile texture on-line checkingi sorter |
| US20170116728A1 (en) * | 2015-10-22 | 2017-04-27 | International Business Machines Corporation | Viewpoint recognition in computer tomography images |
| CN106599848A (en) * | 2016-12-16 | 2017-04-26 | 南京理工大学 | Depth visual feature and support vector machine-based terrain texture recognition algorithm |
| CN107192716A (en) * | 2017-04-26 | 2017-09-22 | 广东工业大学 | A kind of workpiece, defect quick determination method based on contour feature |
Non-Patent Citations (1)
| Title |
|---|
| 冀中 等: "基于卷积神经网络的纹理分类方法研究", 《计算机科学与探索》 * |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109051106A (en) * | 2018-06-28 | 2018-12-21 | 广东科达洁能股份有限公司 | A kind of ceramic tile potted flower color packing method and system |
| CN109051106B (en) * | 2018-06-28 | 2021-08-20 | 广东科达洁能股份有限公司 | Ceramic tile pattern-dividing packaging method and system |
| CN109523505A (en) * | 2018-09-18 | 2019-03-26 | 深圳市智信精密仪器有限公司 | A kind of ceramic brick surface patterned defect detection method based on machine vision |
| CN109523505B (en) * | 2018-09-18 | 2021-06-11 | 深圳市智信精密仪器有限公司 | Method for detecting pattern defects on surface of ceramic tile based on machine vision |
| CN109697303A (en) * | 2018-11-16 | 2019-04-30 | 台州旭日环境科技有限公司 | A kind of building brick sticking scheme generation method towards plane special-shaped region |
| CN109829496B (en) * | 2019-01-30 | 2021-05-25 | 广州市载道信息科技有限公司 | Physical measurement classification method and equipment |
| CN109829496A (en) * | 2019-01-30 | 2019-05-31 | 广州市载道信息科技有限公司 | A kind of physical measurement classification method and equipment |
| CN110930352B (en) * | 2019-02-25 | 2024-02-13 | 研祥智慧物联科技有限公司 | Object color difference defect detection method, device, computer equipment and storage medium |
| CN110930352A (en) * | 2019-02-25 | 2020-03-27 | 研祥智能科技股份有限公司 | Object color difference defect detection method and device, computer equipment and storage medium |
| WO2020173177A1 (en) * | 2019-02-25 | 2020-09-03 | 研祥智能科技股份有限公司 | Object color difference defect detection method, device, computer device, and storage medium |
| CN110163244A (en) * | 2019-04-04 | 2019-08-23 | 睿视智觉（厦门）科技有限公司 | A kind of ceramic tile texture classifying method and device |
| CN110702839B (en) * | 2019-09-17 | 2021-03-02 | 珠海格力电器股份有限公司 | Method and device for detecting gas pollutants |
| CN110702839A (en) * | 2019-09-17 | 2020-01-17 | 珠海格力电器股份有限公司 | Method and device for detecting gas pollutants |
| CN112633393A (en) * | 2020-12-29 | 2021-04-09 | 北京理工大学重庆创新中心 | Automatic classification method and device for ceramic tile textures |
| CN113674305A (en) * | 2021-08-25 | 2021-11-19 | 茶陵县强强陶瓷有限公司 | Intelligent blank storage system applied to ceramic production and control method |
| CN113674305B (en) * | 2021-08-25 | 2022-05-31 | 茶陵县强强陶瓷有限公司 | Intelligent blank storage system applied to ceramic production and control method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107730500A (en) | A kind of ceramic tile texture detection, system, device and readable storage medium storing program for executing | |
| CN111179251B (en) | Defect detection system and method based on twin neural network and by utilizing template comparison | |
| CN106568783B (en) | A kind of hardware defect detecting system and method | |
| CN106875373B (en) | Mobile phone screen MURA defect detection method based on convolutional neural network pruning algorithm | |
| CN108765412B (en) | Strip steel surface defect classification method | |
| CN111402226A (en) | Surface defect detection method based on cascade convolution neural network | |
| CN110853015A (en) | Aluminum profile defect detection method based on improved Faster-RCNN | |
| CN109376792A (en) | Photovoltaic cell open defect classification method based on multichannel residual error neural network | |
| CN110111297A (en) | A kind of injection-molded item surface image defect identification method based on transfer learning | |
| CN110274908A (en) | Flaw detection apparatus, defect detecting method and computer readable recording medium | |
| CN106373124B (en) | Industrial products surface defect visible detection method based on gray level co-occurrence matrixes and RANSAC | |
| CN111260614A (en) | Convolutional neural network cloth flaw detection method based on extreme learning machine | |
| CN108830285A (en) | A kind of object detection method of the reinforcement study based on Faster-RCNN | |
| CN107064170A (en) | One kind detection phone housing profile tolerance defect method | |
| KR20190075707A (en) | Method for sorting products using deep learning | |
| Denhof et al. | Automatic optical surface inspection of wind turbine rotor blades using convolutional neural networks | |
| CN103544499A (en) | Method for reducing dimensions of texture features for surface defect detection on basis of machine vision | |
| CN106650823A (en) | Probability extreme learning machine integration-based foam nickel surface defect classification method | |
| Pawening et al. | Classification of textile image using support vector machine with textural feature | |
| CN107622276B (en) | Deep learning training method based on combination of robot simulation and physical sampling | |
| CN109509171A (en) | A kind of Fabric Defects Inspection detection method based on GMM and image pyramid | |
| CN114549507B (en) | Improved Scaled-YOLOv fabric flaw detection method | |
| Yazdchi et al. | Detection and classification of surface defects of cold rolling mill steel using morphology and neural network | |
| CN116152214A (en) | Plastic powder weather resistance detection method for metal product coating | |
| CN206897873U (en) | A kind of image procossing and detecting system based on detection product performance |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180223 |
|
| RJ01 | Rejection of invention patent application after publication |