CN108564104A - Product defects detection method, device, system, server and storage medium - Google Patents
Product defects detection method, device, system, server and storage medium Download PDFInfo
- Publication number
- CN108564104A CN108564104A CN201810020247.8A CN201810020247A CN108564104A CN 108564104 A CN108564104 A CN 108564104A CN 201810020247 A CN201810020247 A CN 201810020247A CN 108564104 A CN108564104 A CN 108564104A
- Authority
- CN
- China
- Prior art keywords
- classification
- defect
- prediction result
- server
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- 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]
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Health & Medical Sciences (AREA)
- Quality & Reliability (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
A kind of product defects detection method of present invention proposition, device, system, server and computer readable storage medium, wherein product defects detection method include:Obtain the image data of product;Convert described image data to classification request, execute server is determined according to the deployment scenario of the classification prediction model on multiple servers, and classification request is sent to the execute server, to provide the prediction result of defect classification by the classification prediction model in the execute server.Embodiment provided by the invention can make model can adapt to the newest demand of production environment with business development iterative model, bring and be obviously improved for industrial production line in nicety of grading, scalability, standardization etc., and parallel processing further improves efficiency.
Description
Technical field
The present invention relates to information technology field more particularly to a kind of product defects detection method, device, system, servers
And computer readable storage medium.
Background technology
The quality inspection links in many production industries mainly carry out defect with visual manner for product surface image at present
Detection.By taking paper-making industry as an example, the detection of paper quality is mainly judged according to the picture of paper surface.In paper-making industry production line
On, quality inspection is mostly that manual inspection or semi-automatic optical instrument assist quality inspection, not only inefficiency, but also is susceptible to mistake
Sentence;In addition, the industrial data that this mode generates is not easy to store, manage and mining again recycles.The manual inspection the case where
Under, need business expert to carry out walkaround inspection in production scene, manual record gets off to do subsequent processing again after finding defect.This
Not only efficiency is low for kind method, is easy erroneous judgement of failing to judge, and data are difficult to carry out secondary use excavation, and production environment often compares evil
It is bad, the health and safety of personnel can be adversely affected.Latter quality inspection mode is mostly to be based on traditional expert system or feature
The quality inspection system of engineering, feature and decision rule are all based on experience and are cured in machine, it is difficult to the development iteration of business,
Lead to the development with production technology, the accuracy of detection of system is lower and lower, or even is reduced to complete unusable state.This
Outside, the feature of traditional quality inspection system is all cured by third-party vendor within hardware in advance, and when upgrading not only needs to production line
Key technological transformation is carried out, and expensive.Traditional quality inspection system safety, standardization, scalability etc. all there is
It is apparent insufficient, it is unfavorable for the optimization and upgrading of production line.
Invention content
A kind of product defects detection method of offer of the embodiment of the present invention, device, system, server and computer-readable storage
Medium, at least to solve one or more technical problems in the prior art.
In a first aspect, an embodiment of the present invention provides a kind of product defects detection methods, including:Obtain the picture number of product
According to;Convert described image data to classification request, the deployment scenario according to the classification prediction model on multiple servers is true
Determine execute server, and classification request is sent to execute server, with pre- by the classification in the execute server
Survey the prediction result that model provides defect classification.
With reference to first aspect, the present invention is in the first embodiment of first aspect, according on multiple servers
The deployment scenario of classification prediction model determines execute server, including:Pre-set server resource distribution management table is inquired,
The server resource distribution management table is used to record the load condition of each server for being deployed with the classification prediction model;
The load condition of more each server for being deployed with the classification prediction model, the server for loading minimum is determined as executing
Server.
With reference to first aspect, the first embodiment of first aspect, second embodiment party of the present invention in first aspect
In formula, the prediction result for the defect classification that the execute server returns is received;It is done according to the prediction result of the defect classification
Go out corresponding defect processing operation, the defect processing operation includes:It alarms, label, storing daily record and/or shutdown.
The first embodiment with reference to first aspect makes corresponding defect according to the prediction result of the defect classification
Processing operation, including:The corresponding pass operated with the defect processing according to the prediction result of the pre-set defect classification
System makes corresponding defect processing operation;Alternatively, according to the grade of the prediction result of the pre-set defect classification, with
And the correspondence of the grade of the prediction result of the defect classification and defect processing operation, make corresponding defect processing
Operation.
Second aspect, an embodiment of the present invention provides a kind of product defects detection methods, including:Receive the picture number of product
According to classification request;The classification is asked by classification prediction model trained in advance to carry out classified calculating, provides defect class
Other prediction result;The classification prediction model includes Feature Selection Model and defect location disaggregated model;The feature extraction
Model is used to extract the feature of the image data in the classification request;The defect location disaggregated model is used for according to extraction
The feature of described image data provides the prediction result of defect classification, and the prediction result includes:Figure in the classification request
Seem the classification and position coordinates of no existing defects and defect.
In conjunction with second aspect, the present invention is asking to carry out in the first embodiment of second aspect to the classification
Before classified calculating, further include:Image data in the classification request is pre-processed, the pretreatment includes that image is gone
It makes an uproar, remove background, compression of images and/or format conversion.
In conjunction with the first embodiment of second aspect, second aspect, second embodiment party of the present invention in second aspect
In formula, further include:According to the history labeled data of the image data of product, training obtains the classification prediction model in advance.
In conjunction with the third embodiment of second aspect, the present invention is described in the third embodiment of second aspect
Feature Selection Model includes depth convolutional neural networks;The defect location disaggregated model includes RCNN, SSD or Mask RCNN.
In conjunction with the first embodiment of second aspect, second aspect, four kind embodiment party of the present invention in second aspect
In formula, after the prediction result for providing defect classification, further include:It is made according to the prediction result of the defect classification corresponding
Defect processing operation, defect processing operation includes:It alarms, label, storing daily record and/or shutdown.
In conjunction with the 4th kind of embodiment of second aspect, corresponding defect is made according to the prediction result of the defect classification
Processing operation, including:The corresponding pass operated with the defect processing according to the prediction result of the pre-set defect classification
System makes corresponding defect processing operation;Alternatively, according to the grade of the prediction result of the pre-set defect classification, with
And the correspondence of the grade of the prediction result of the defect classification and defect processing operation, make corresponding defect processing
Operation.
The third aspect, an embodiment of the present invention provides a kind of product defects detection devices, including:Data acquisition module is used
In the image data for obtaining product;Load balancing module, for converting described image data to classification request, according to multiple
The deployment scenario of classification prediction model on server determines execute server, and classification request is sent to the service of execution
Device, to provide the prediction result of defect classification by the classification prediction model in the execute server.
In conjunction with the third aspect, the present invention in the first embodiment of the third aspect, also use by the load balancing module
In:Pre-set server resource distribution management table is inquired, the server resource distribution management table is for recording each portion
There is the load condition of the server of the classification prediction model in administration;More each server for being deployed with the classification prediction model
Load condition, the server for loading minimum is determined as execute server.
Fourth aspect, an embodiment of the present invention provides a kind of product defects detection devices, including:Data reception module is used
In the classification request for the image data for receiving product;Classification prediction model, for passing through classification prediction model pair trained in advance
The classification request carries out classified calculating, provides the prediction result of defect classification;The classification prediction model includes feature extraction
Model and defect location disaggregated model;The Feature Selection Model is used to extract the spy of the image data in the classification request
Sign;The defect location disaggregated model is used to provide the prediction knot of defect classification according to the feature of the described image data of extraction
Fruit, the prediction result include:Image in the classification request is sat with the presence or absence of the classification and position of defect and defect
Mark.
In conjunction with fourth aspect, the present invention further includes control module, is used in the first embodiment of fourth aspect:
Corresponding defect processing operation is made according to the prediction result of the defect classification, the defect processing operation includes:Alarm,
It labels, store daily record and/or shutdown.
In a possible design, the structure of product defects detection device includes processor and memory, described to deposit
Reservoir supports product defects detection device to execute product defects detection method in above-mentioned first aspect or second aspect for storing
Program, the processor is configurable for executing the program stored in the memory.
5th aspect, an embodiment of the present invention provides a kind of product defects detecting system, including the above-mentioned third aspect or the
Any device in four aspects, and, Production database, the image data for storing product, and with the product
The corresponding defect classification of image data prediction result and defect processing corresponding with the prediction result of defect classification behaviour
Make;Tranining database, the history labeled data of the image data for storing product, the history labeled data is for training point
Class prediction model.
6th aspect, an embodiment of the present invention provides a kind of servers, including:One or more processors;Storage device,
For storing one or more programs;When one or more of programs are executed by one or more of processors so that
One or more of processors realize the method as described in any in above-mentioned first aspect or second aspect.
7th aspect, an embodiment of the present invention provides a kind of computer readable storage mediums, are stored with computer program,
The program realizes any method in above-mentioned first aspect or second aspect when being executed by processor.
A technical solution in above-mentioned technical proposal has the following advantages that or advantageous effect:Embodiment provided by the invention
Suitable for any scene for carrying out defect classification using human eye, photo or machine vision, mould can be made with business development iterative model
Type can adapt to the newest demand of production environment, be brought for industrial production line in nicety of grading, scalability, standardization etc.
It is obviously improved.
Another technical solution in above-mentioned technical proposal has the following advantages that or advantageous effect:Pass through load balancing and tune
Degree, parallel processing further improve working efficiency.
Above-mentioned general introduction is merely to illustrate that the purpose of book, it is not intended to be limited in any way.Except foregoing description
Schematical aspect, except embodiment and feature, by reference to attached drawing and the following detailed description, the present invention is further
Aspect, embodiment and feature, which will be, to be readily apparent that.
Description of the drawings
In the accompanying drawings, unless specified otherwise herein, otherwise run through the identical reference numeral of multiple attached drawings and indicate same or analogous
Component or element.What these attached drawings were not necessarily to scale.It should be understood that these attached drawings are depicted only according to the present invention
Some disclosed embodiments, and should not serve to limit the scope of the present invention.
Fig. 1 is the general frame figure of the product defects detection method of the embodiment of the present invention;
Fig. 2 is a kind of step flow chart of preferred embodiment of product defects detection method provided by the invention;
Fig. 3 is a kind of work of preferred embodiment at the execute server end of product defects detection method provided by the invention
Flow chart;
Fig. 4 is a kind of workflow schematic diagram of preferred embodiment of product defects detection method provided by the invention;
Fig. 5 is a kind of principle schematic of preferred embodiment of product defects detection method provided by the invention;
Fig. 6 is the general frame figure of the product defects detection device of the embodiment of the present invention;
Fig. 7 is a kind of structure of preferred embodiment at the execute server end of product defects detection device provided by the invention
Schematic diagram;
Fig. 8 is the structure of another preferred embodiment at the execute server end of product defects detection device provided by the invention
Schematic diagram.
Specific implementation mode
Hereinafter, certain exemplary embodiments are simply just described.As one skilled in the art will recognize that
Like that, without departing from the spirit or scope of the present invention, described embodiment can be changed by various different modes.
Therefore, attached drawing and description are considered essentially illustrative rather than restrictive.
An embodiment of the present invention provides a kind of product defects detection methods.Fig. 1 is that the product defects of the embodiment of the present invention are examined
The general frame figure of survey method.As shown in Figure 1, the product defects detection method of the embodiment of the present invention includes:Step S110 is obtained
The image data of product;Step S120 converts described image data to classification request, according to the classification on multiple servers
The deployment scenario of prediction model determines execute server, and classification request is sent to execute server, described to pass through
Classification prediction model in execute server provides the prediction result of defect classification.
There are mainly two types of modes in defect classification application for existing quality inspection system.First is pure artificial quality inspection mode, i.e.,
The photo in production environment, which is visually observed, dependent on industry specialists provides judgement;The second artificial quality inspection mode assisted for machine,
It is mainly filtered out by the quality inspection system with certain judgement and does not have defective photo, by industry specialists to doubtful existing defects
Photo be detected judgement.Wherein, the second way is mostly expert system and Feature Engineering System Development, and expert will be through
It tests and is solidificated in quality inspection system.In both the above method, defects detection and localization method depend critically upon expertise, and there are standards
Really rate is low, real-time is poor, is difficult to extend, is not easy to evolve with business, standardize the shortcomings of weak.
Application of the embodiment of the present invention based on artificial intelligence technology in machine vision, is being produced using image capture device
The image acquired in real time on line is detected judgement by advance trained machine learning model to the surface quality of product,
If detected on the current product by image capture device there are quality problems, the class corresponding to the quality problems is judged
Not.The embodiment of the present invention is suitable for the scene of the image data detection product quality of the with good grounds product surface of institute, compared to existing skill
For art dependent on artificial and expertise detection method, high degree of automation, accuracy of detection be high, can be with business development iteration mould
Type, scalability, standardization etc. are obviously improved, and by load balancing and scheduling, parallel processing further improves work
Make efficiency.
Fig. 2 is a kind of step flow chart of preferred embodiment of product defects detection method provided by the invention.Such as Fig. 2 institutes
Show, according to a kind of embodiment of product defects detection method of the present invention, according to the classification prediction model on multiple servers
Deployment scenario determine execute server, including:Step S210 inquires pre-set server resource distribution management table, institute
State load condition of the server resource distribution management table for recording each server for being deployed with the classification prediction model;Step
Rapid S220, the load condition of more each server for being deployed with the classification prediction model are true by the server for loading minimum
It is set to execute server.
In this embodiment, the image data of product is obtained, that is, obtains the picture generated in real time on production line,
Then it converts the image data of product to classification request (query), and classification request is sent to execute server.Its
In, execute server is the determination according to deployment scenario of the classification prediction model on multiple servers, in advance by a version
This classification prediction model copies to does parallel processing on multiple machines, and the deployment scenario according to prediction model of classifying on line is real-time
Carry out load balancing and scheduling, by classification request be sent to best carrying classify prediction model server on.It can be arranged
Machine resources distribution management table, the load condition for recording each machine in real time select a load at least, i.e. to respond most fast
Classification request is sent to the server by server.
According to a kind of embodiment of product defects detection method of the present invention, the prediction result for providing defect classification it
Afterwards, further include:Corresponding defect processing operation, the defect processing operation are made according to the prediction result of the defect classification
Including:It alarms, label, storing daily record and/or shutdown.
According to a kind of embodiment of product defects detection method of the present invention, done according to the prediction result of the defect classification
Go out corresponding defect processing operation, including:According to the prediction result of the pre-set defect classification and the defect processing
The correspondence of operation makes corresponding defect processing operation;Alternatively, according to the prediction knot of the pre-set defect classification
The correspondence of the grade of the prediction result of the grade of fruit and the defect classification and defect processing operation, makes phase
The defect processing operation answered.
On the other hand, an embodiment of the present invention provides a kind of product defects detection methods.Fig. 3 is product provided by the invention
A kind of work flow diagram of the preferred embodiment at the execute server end of defect inspection method.As shown in figure 3, the embodiment of the present invention
The product defects detection method at execute server end include:Step S310 receives the classification request of the image data of product;Step
Rapid S320 asks the classification by classification prediction model trained in advance to carry out classified calculating, provides the pre- of defect classification
Survey result.
According to a kind of embodiment of product defects detection method of the present invention, the classification is being asked to carry out classified calculating
Before, further include:Image data in the classification request is pre-processed, the pretreatment includes image denoising, removal
Background, compression of images and/or format conversion.To pre-processing for image data, effectively will partly retain follow-up to carry out
Processing.
According to a kind of embodiment of product defects detection method of the present invention, further include:According to the image data of product
Training obtains the classification prediction model to history labeled data in advance.Fig. 4 is product defects detection method provided by the invention
A kind of workflow schematic diagram of preferred embodiment.As shown in figure 4, classification prediction model is marked according to history by training engine
What data were trained, history labeled data is stored in tranining database, and training engine is asked to tranining database transmission data
It asks, training data is returned to trained engine by the request of tranining database response data.In addition, be stored in Production database including
The image data of recent product, and the number including the prediction result of defect classification corresponding with the image data of the product
According to, Production database can be that tranining database provides data update at any time, if production technology development has updated, training data
Training data in library can make model can adapt to the newest demand of production environment with the development iteration of business.
According to a kind of embodiment of product defects detection method of the present invention, the classification prediction model includes feature extraction
Model and defect location disaggregated model;The Feature Selection Model is used to extract the spy of the image data in the classification request
Sign;The defect location disaggregated model is used to provide the prediction knot of defect classification according to the feature of the described image data of extraction
Fruit, the prediction result include:Image in the classification request is sat with the presence or absence of the classification and position of defect and defect
Mark.
According to a kind of embodiment of product defects detection method of the present invention, the Feature Selection Model includes depth convolution
Neural network;The defect location disaggregated model includes RCNN (Region Based CNN, region convolutional neural networks), SSD
(Single Shot MultiBoxDetector) or Mask RCNN.
Fig. 5 is a kind of principle schematic of preferred embodiment of product defects detection method provided by the invention.Such as Fig. 5 institutes
Show, in one embodiment, using depth convolutional neural networks as Feature Selection Model, with object detection, image segmentation
Technology is as defect location sorter network, input of the original image as model on production line, to the image data of input
After being pre-processed, neural network model effectively will be partly inputted, depth convolutional neural networks extract the spy in original image
Sign, and input the feature into defect location sorter network.RCNN, SSD, Mask RCNN can be used in defect location sorter network
Etc. object detections and other Image Segmentation Models, as defect location disaggregated model, according to the extraction of depth convolutional neural networks
Feature judges that a certain position in picture then judges the classification belonging to defect with the presence or absence of defect if there is defect.Model
Final output be picture present in defect classification and its relative position coordinates in picture.If existed in picture more
A defect, model can provide the classification and its relative coordinate of each defect.By taking paper production industry as an example, the classification packet of defect
Include fold, broken hole, tearing, impurity and aberration etc..
Trained model each time can gradually be replaced in such a way that small flow is reached the standard grade just running on line it is old
Model is extended with to achieve the purpose that model with service dynamic extensive.
In another embodiment, depth convolutional neural networks CNN (Convolutional Neural are also based on
Network, convolutional neural networks) except machine learning method do feature extraction network, may be used except object detection and
Model other than image segmentation is as defect location sorter network model.
According to a kind of embodiment of product defects detection method of the present invention, the prediction result for providing defect classification it
Afterwards, further include:Corresponding defect processing operation, the defect processing operation are made according to the prediction result of the defect classification
Including:It alarms, label, storing daily record and/or shutdown.
Referring to Fig. 4, after providing the prediction result for representing the defect classification, prediction result is sent to control module,
Control module is combined design with business scenario, can be made to the prediction result that model provides according to business demand and meet production
The response that environment scene requires, such as alarm, label, storing daily record and/or shutdown, control module is by prediction result and response
Processing behavior as producing daily record storage on line in Production database.
According to a kind of embodiment of product defects detection method of the present invention, done according to the prediction result of the defect classification
Go out corresponding defect processing operation, including:According to the prediction result of the pre-set defect classification and the defect processing
The correspondence of operation makes corresponding defect processing operation;Alternatively, according to the prediction knot of the pre-set defect classification
The correspondence of the grade of the prediction result of the grade of fruit and the defect classification and defect processing operation, makes phase
The defect processing operation answered.
Wherein, the prediction result of the prediction result of defect classification is operated with defect processing correspondence, defect classification
The correspondence of the grade of the prediction result of grade and defect classification and defect processing operation, can be all arranged by system default
Or by the self-defined setting of manufacturer.For example, the grade of the prediction result of defect classification can be divided into it is serious, general or not tight
Weight;Shutdown operation does in producer when the serious disfigurement discovery of grade can be arranged, when grade general disfigurement discovery does alarm operation, etc.
Grade not serious disfigurement discovery when only label operation etc..
On the other hand, an embodiment of the present invention provides a kind of product defects detection devices.Fig. 6 is the production of the embodiment of the present invention
The general frame figure of product defect detecting device.As shown in fig. 6, the product defects detection device of the embodiment of the present invention includes:Data
Acquisition module 100, the image data for obtaining product;Load balancing module 200, for converting point described image data to
Class is asked, and execute server is determined according to the deployment scenario of the classification prediction model on multiple servers, and by the classification
Request is sent to execute server, to provide the prediction knot of defect classification by the classification prediction model in the execute server
Fruit.
According to a kind of embodiment of product defects detection device of the present invention, the load balancing module 200 is additionally operable to:It looks into
Pre-set server resource distribution management table is ask, the server resource distribution management table is for having recorded each deployment
State the load condition of the server of classification prediction model;The load of more each server for being deployed with the classification prediction model
The server for loading minimum is determined as execute server by state.
It is (i.e. negative comprising prediction engine Predictor in the product defects detection device of the embodiment of the present invention referring back to Fig. 4
Carry balance module), classification prediction model Classifier, training engine Trainer.Wherein, prediction engine will be real on production line
When the picture that generates be converted into classification request (query), and carry out loading in real time according to the deployment scenario of prediction model on line
Classification request is sent on the best server for carrying prediction model by weighing apparatus and scheduling.When being run really on the server
Classification prediction model, classification prediction model are trained to obtain by training engine according to history labeled data.
According to a kind of embodiment of product defects detection device of the present invention, further include:The execute server is received to return
The prediction result for the defect classification returned;Corresponding defect processing operation, institute are made according to the prediction result of the defect classification
Stating defect processing operation includes:It alarms, label, storing daily record and/or shutdown.
According to a kind of embodiment of product defects detection device of the present invention, done according to the prediction result of the defect classification
Go out corresponding defect processing operation, including:According to the prediction result of the pre-set defect classification and the defect processing
The correspondence of operation makes corresponding defect processing operation;Alternatively, according to the prediction knot of the pre-set defect classification
The correspondence of the grade of the prediction result of the grade of fruit and the defect classification and defect processing operation, makes phase
The defect processing operation answered.
In another aspect, an embodiment of the present invention provides a kind of product defects detection devices.Fig. 7 is product provided by the invention
A kind of structural schematic diagram of the preferred embodiment at the execute server end of defect detecting device.As shown in fig. 7, the embodiment of the present invention
The product defects detection device at execute server end include:Data reception module 300, the image data for receiving product
Classification request;Classification prediction model 400, for being classified to classification request by classification prediction model trained in advance
It calculates, provides the prediction result of defect classification.
According to a kind of embodiment of product defects detection device of the present invention, the classification prediction model 400 is additionally operable to:
Before asking to carry out classified calculating to the classification, the image data in the classification request is pre-processed, the pre- place
Reason includes image denoising, removal background, compression of images and/or format conversion.
Further include trained engine according to a kind of embodiment of product defects detection device of the present invention, for according to product
Image data history labeled data in advance training obtain the classification prediction model.
According to a kind of embodiment of product defects detection device of the present invention, the classification prediction model 400 includes feature
Extraction model and defect location disaggregated model;The Feature Selection Model is used to extract the image data in the classification request
Feature;The defect location disaggregated model is used to provide the prediction knot of defect classification according to the feature of the described image data of extraction
Fruit, the prediction result include:Image in the classification request is sat with the presence or absence of the classification and position of defect and defect
Mark.
According to a kind of embodiment of product defects detection device of the present invention, the Feature Selection Model includes depth convolution
Neural network;The defect location disaggregated model includes RCNN, SSD or Mask RCNN.
Fig. 8 is the structure of another preferred embodiment at the execute server end of product defects detection device provided by the invention
Schematic diagram.Further include control module 600 as shown in figure 8, according to a kind of embodiment of product defects detection device of the present invention,
For:Corresponding defect processing operation is made according to the prediction result of the defect classification, the defect processing operation includes:
It alarms, label, storing daily record and/or shutdown.
According to a kind of embodiment of product defects detection device of the present invention, the control module 600 is additionally operable to:According to pre-
The correspondence of the prediction result for the defect classification being first arranged and defect processing operation, makes corresponding defect processing
Operation;Alternatively, according to the prediction knot of the grade of the prediction result of the pre-set defect classification and the defect classification
The correspondence of the grade of fruit and defect processing operation makes corresponding defect processing operation.
Referring back to Fig. 4, the product defects detection device of the embodiment of the present invention include mainly prediction engine Predictor (i.e.
Load balancing module), classification prediction model Classifier, training engine Trainer, control module Controller, data
The several main modulars of library Database.
Wherein, prediction engine converts the picture generated in real time on production line to classification request (query), and according on line
The deployment scenario of prediction model carries out load balancing and scheduling in real time, and classification request is sent to and best carries prediction model
Server on.Real-time grading prediction model is run on the server, which completes via training engine training.Model
After carrying out preset pretreatment for the image data in the classification request of arrival, classified calculating is carried out, and is provided and is represented this and lack
The prediction result of classification is fallen into, and result is sent to control module.Control module is combined design with business scenario, can be according to industry
Business demand makes the prediction result that model provides the response for meeting production environment scene requirement, such as alarm, storage daily record.
Control module can be using the processing behavior of prediction result and response as in production daily record storage to Production database on line.Classification is pre-
Model is surveyed to train to obtain according to history labeled data by training engine.
In a possible design, the structure of product defects detection device includes processor and memory, described to deposit
Reservoir supports product defects detection device to execute product defects detection method in above-mentioned first aspect or second aspect for storing
Program, the processor is configurable for executing the program stored in the memory.
Another aspect, an embodiment of the present invention provides a kind of product defects detecting systems, including the above-mentioned third aspect or
Any device in four aspects, and, Production database, the image data for storing product, and with the product
The corresponding defect classification of image data prediction result and defect processing corresponding with the prediction result of defect classification behaviour
Make;Tranining database, the history labeled data of the image data for storing product, the history labeled data is for training point
Class prediction model.Referring to Fig. 4, history labeled data is stored in tranining database, and training engine sends number to tranining database
According to request, training data is returned to trained engine by the request of tranining database response data.In addition, being stored in Production database
The image data of recent product, and defect classification corresponding with the image data of the product prediction result, creation data
Library can be that tranining database provides data update at any time, if production technology development has updated, the training in tranining database
Data can make model can adapt to the newest demand of production environment with the development iteration of business.
In another aspect, an embodiment of the present invention provides a kind of servers, including:One or more processors;Storage device,
For storing one or more programs;When one or more of programs are executed by one or more of processors so that
One or more of processors realize the method as described in any in above-mentioned first aspect or second aspect.
In another aspect, an embodiment of the present invention provides a kind of computer readable storage medium, it is stored with computer program,
The program realizes any method in above-mentioned first aspect or second aspect when being executed by processor.
A technical solution in above-mentioned technical proposal has the following advantages that or advantageous effect:Embodiment provided by the invention
Suitable for any scene for carrying out defect classification using human eye, photo or machine vision, mould can be made with business development iterative model
Type can adapt to the newest demand of production environment, be brought for industrial production line in nicety of grading, scalability, standardization etc.
It is obviously improved.
Another technical solution in above-mentioned technical proposal has the following advantages that or advantageous effect:Pass through load balancing and tune
Degree, parallel processing further improve working efficiency.
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 included at least one embodiment or example of the invention.Moreover, particular features, structures, materials, or characteristics described
It may be combined in any suitable manner in any one or more of the embodiments or examples.In addition, without conflicting with each other, this
The technical staff in field can be by the spy of different embodiments or examples described in this specification and different embodiments or examples
Sign is combined.
In addition, term " first ", " second " are used for description purposes only, it is not understood to indicate or imply relative importance
Or implicitly indicate the quantity of indicated technical characteristic." first " is defined as a result, the feature of " second " can be expressed or hidden
Include at least one this feature containing ground.In the description of the present invention, the meaning of " plurality " is two or more, unless otherwise
Clear specific restriction.
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 specific logical function or process the step of executable instruction code module, segment or portion
Point, and the range of the preferred embodiment of the present invention includes other realization, wherein can not press shown or discuss suitable
Sequence, include according to involved function by it is basic simultaneously in the way of or in the opposite order, to execute function, this should be of the invention
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 (system of such as computer based system including 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 " can any can be included, store, communicating, propagating or passing
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 includes following:Electricity with one or more wiring
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 read-only memory
(CDROM).In addition, computer-readable medium can even is that the paper that can print described program on it or other suitable Jie
Matter, because can be for example by carrying out optical scanner to paper or other media, then into edlin, interpretation or when necessary with other
Suitable method is handled electronically to obtain described program, is then stored in computer storage.
It should be appreciated that each section of the present invention can be realized with hardware, software, firmware or combination thereof.Above-mentioned
In embodiment, software that multiple steps or method can in memory and by suitable instruction execution system be executed with storage
Or firmware is realized.It, and in another embodiment, can be under well known in the art for example, if realized with hardware
Any one of row technology or their combination are realized:With the logic gates for realizing logic function to data-signal
Discrete logic, with suitable combinational logic gate circuit application-specific integrated circuit, programmable gate array (PGA), scene
Programmable gate array (FPGA) etc..
Those skilled in the art are appreciated that realize all or part of step that above-described embodiment method carries
Suddenly it is that relevant hardware can be instructed to complete by program, the program can be stored in a kind of computer-readable storage medium
In matter, which includes the steps that one or a combination set of embodiment of the method when being executed.Wherein device embodiments and method
Embodiment is corresponding, therefore the embodiment description of device is simpler, and associated description can refer to the embodiment of method
Description.
In addition, each functional unit in each embodiment of the present invention can be integrated in a processing module, it can also
That each unit physically exists alone, can also two or more units be integrated in a module.Above-mentioned integrated mould
The form that hardware had both may be used in block is realized, can also be realized in the form of software function module.The integrated module is such as
Fruit is realized in the form of software function module and when sold or used as an independent product, can also be stored in a computer
In readable storage medium storing program for executing.The storage medium can be read-only memory, disk or CD etc..
The above description is merely a specific embodiment, but scope of protection of the present invention is not limited thereto, any
Those familiar with the art in the technical scope disclosed by the present invention, can readily occur in its various change or replacement,
These should be covered by the protection scope of the present invention.Therefore, protection scope of the present invention should be with the guarantor of the claim
It protects subject to range.
Claims (17)
1. a kind of product defects detection method, which is characterized in that including:
Obtain the image data of product;
Convert described image data to classification request, the deployment scenario according to the classification prediction model on multiple servers is true
Determine execute server, and classification request is sent to the execute server, to pass through point in the execute server
Class prediction model provides the prediction result of defect classification.
2. according to the method described in claim 1, it is characterized in that, according to the portion of the classification prediction model on multiple servers
Administration's situation determines execute server, including:
Pre-set server resource distribution management table is inquired, the server resource distribution management table is for recording each portion
There is the load condition of the server of the classification prediction model in administration;
The load condition of more each server for being deployed with the classification prediction model, the server for loading minimum is determined as
Execute server.
3. method according to claim 1 or 2, which is characterized in that further include:
Receive the prediction result for the defect classification that the execute server returns;
Corresponding defect processing operation is made according to the prediction result of the defect classification, the defect processing operation includes:
It alarms, label, storing daily record and/or shutdown.
4. according to the method described in claim 3, it is characterized in that, being made accordingly according to the prediction result of the defect classification
Defect processing operates, including:
According to the correspondence of the prediction result of the pre-set defect classification and defect processing operation, make corresponding
Defect processing operation;Alternatively,
According to the prediction result of the grade of the prediction result of the pre-set defect classification and the defect classification etc.
The correspondence of grade and defect processing operation makes corresponding defect processing operation.
5. a kind of product defects detection method, which is characterized in that including:
Receive the classification request of the image data of product;
The classification is asked by classification prediction model trained in advance to carry out classified calculating, provides the prediction knot of defect classification
Fruit;The classification prediction model includes Feature Selection Model and defect location disaggregated model;The Feature Selection Model is for carrying
Take the feature of the image data in the classification request;The defect location disaggregated model is used for the described image number according to extraction
According to feature provide the prediction result of defect classification, the prediction result includes:Image in the classification request whether there is
The classification and position coordinates of defect and defect.
6. according to the method described in claim 5, it is characterized in that, before asking to carry out classified calculating to the classification, go back
Including:Image data in the classification request is pre-processed, the pretreatment includes image denoising, removal background, figure
As compression and/or format conversion.
7. method according to claim 5 or 6, which is characterized in that further include:According to the history mark of the image data of product
Training obtains the classification prediction model to note data in advance.
8. method according to claim 5 or 6, which is characterized in that the Feature Selection Model includes depth convolutional Neural
Network;The defect location disaggregated model includes RCNN, SSD or Mask RCNN.
9. method according to claim 5 or 6, which is characterized in that after the prediction result for providing defect classification, also wrap
It includes:
Corresponding defect processing operation is made according to the prediction result of the defect classification, the defect processing operation includes:
It alarms, label, storing daily record and/or shutdown.
10. according to the method described in claim 9, it is characterized in that, being made accordingly according to the prediction result of the defect classification
Defect processing operation, including:
According to the correspondence of the prediction result of the pre-set defect classification and defect processing operation, make corresponding
Defect processing operation;Alternatively,
According to the prediction result of the grade of the prediction result of the pre-set defect classification and the defect classification etc.
The correspondence of grade and defect processing operation makes corresponding defect processing operation.
11. a kind of product defects detection device, which is characterized in that including:
Data acquisition module, the image data for obtaining product;
Load balancing module, it is pre- according to the classification on multiple servers for converting described image data to classification request
The deployment scenario for surveying model determines execute server, and classification request is sent to execute server, to pass through described hold
Classification prediction model on row server provides the prediction result of defect classification.
12. according to the devices described in claim 11, which is characterized in that the load balancing module is additionally operable to:
Pre-set server resource distribution management table is inquired, the server resource distribution management table is for recording each portion
There is the load condition of the server of the classification prediction model in administration;
The load condition of more each server for being deployed with the classification prediction model, the server for loading minimum is determined as
Execute server.
13. a kind of product defects detection device, which is characterized in that including:
Data reception module, the classification request of the image data for receiving product;
Prediction model of classifying is given for carrying out classified calculating to classification request by classification prediction model trained in advance
Go out the prediction result of defect classification;The classification prediction model includes Feature Selection Model and defect location disaggregated model;It is described
Feature Selection Model is used to extract the feature of the image data in the classification request;The defect location disaggregated model is used for root
The prediction result of defect classification is provided according to the feature of the described image data of extraction, the prediction result includes:The classification is asked
Image in asking is with the presence or absence of defect and the classification and position coordinates of defect.
14. device according to claim 13, which is characterized in that further include control module, be used for:According to the defect class
Other prediction result makes corresponding defect processing operation, and the defect processing operation includes:It alarms, label, store day
Will and/or shutdown.
15. a kind of product defects detecting system, which is characterized in that include the device as described in any one of claim 11 or 12
With the device as described in any one of claim 13 or 14, and
Production database, the image data for storing product, and defect classification corresponding with the image data of the product
Prediction result and defect processing corresponding with the prediction result of defect classification operation;
Tranining database, the history labeled data of the image data for storing product, the history labeled data is for training
Classification prediction model.
16. a kind of server, which is characterized in that including:
One or more processors;
Storage device, for storing one or more programs;
When one or more of programs are executed by one or more of processors so that one or more of processors
Realize the method as described in any in claim 1-4 or 5-10.
17. a kind of computer readable storage medium, is stored with computer program, which is characterized in that the program is held by processor
The method as described in any in claim 1-4 or 5-10 is realized when row.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810020247.8A CN108564104A (en) | 2018-01-09 | 2018-01-09 | Product defects detection method, device, system, server and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810020247.8A CN108564104A (en) | 2018-01-09 | 2018-01-09 | Product defects detection method, device, system, server and storage medium |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108564104A true CN108564104A (en) | 2018-09-21 |
Family
ID=63530699
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810020247.8A Pending CN108564104A (en) | 2018-01-09 | 2018-01-09 | Product defects detection method, device, system, server and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108564104A (en) |
Cited By (18)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109472769A (en) * | 2018-09-26 | 2019-03-15 | 成都数之联科技有限公司 | A kind of bad image defect detection method and system |
CN109470708A (en) * | 2018-11-30 | 2019-03-15 | 北京百度网讯科技有限公司 | Quality determining method, device, server and the storage medium of plastic foam cutlery box |
CN109509172A (en) * | 2018-09-25 | 2019-03-22 | 无锡动视宫原科技有限公司 | A kind of liquid crystal display flaw detection method and system based on deep learning |
CN109544522A (en) * | 2018-11-12 | 2019-03-29 | 北京科技大学 | A kind of Surface Defects in Steel Plate detection method and system |
CN109584235A (en) * | 2018-11-30 | 2019-04-05 | 北京百度网讯科技有限公司 | Quality determining method, device, server and the storage medium of plastic foam cutlery box |
CN109598712A (en) * | 2018-11-30 | 2019-04-09 | 北京百度网讯科技有限公司 | Quality determining method, device, server and the storage medium of plastic foam cutlery box |
CN109961433A (en) * | 2019-03-29 | 2019-07-02 | 北京百度网讯科技有限公司 | Product defects detection method, device and computer equipment |
CN110554047A (en) * | 2019-09-06 | 2019-12-10 | 腾讯科技(深圳)有限公司 | method, device, system and equipment for processing product defect detection data |
CN110738237A (en) * | 2019-09-16 | 2020-01-31 | 深圳新视智科技术有限公司 | Defect classification method and device, computer equipment and storage medium |
CN111060512A (en) * | 2018-10-16 | 2020-04-24 | 希捷科技有限公司 | Intelligent defect identification system |
CN111855668A (en) * | 2020-07-17 | 2020-10-30 | 上海洪朴信息科技有限公司 | Product defect detecting system |
CN112070246A (en) * | 2020-08-26 | 2020-12-11 | 上海洪朴信息科技有限公司 | Photovoltaic product defect detection algorithm deployment system |
CN112464495A (en) * | 2020-12-14 | 2021-03-09 | 广东省科学院智能制造研究所 | Defect detection method and device for mold structure |
CN112927170A (en) * | 2021-04-08 | 2021-06-08 | 上海哥瑞利软件股份有限公司 | Automatic defect removal method in semiconductor manufacturing process |
WO2021120186A1 (en) * | 2019-12-20 | 2021-06-24 | 京东方科技集团股份有限公司 | Distributed product defect analysis system and method, and computer-readable storage medium |
CN113129284A (en) * | 2021-02-26 | 2021-07-16 | 山东大学 | Appearance detection method based on 5G cloud edge cooperation and implementation system |
CN113593142A (en) * | 2021-07-26 | 2021-11-02 | 中国工商银行股份有限公司 | Automatic Teller Machine (ATM) patrolling method and device |
US11719727B1 (en) | 2020-02-10 | 2023-08-08 | National Technology & Engineering Solutions Of Sandia, Llc | Systems and methods for screening particle source manufacturing and development test data |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20030202703A1 (en) * | 2002-04-25 | 2003-10-30 | Dainippon Screen Mfg. Co., Ltd. | Apparatus and computer-readable medium for assisting image classification |
CN101996405A (en) * | 2010-08-30 | 2011-03-30 | 中国科学院计算技术研究所 | Method and device for rapidly detecting and classifying defects of glass image |
CN102253049A (en) * | 2011-06-30 | 2011-11-23 | 东北大学 | Method for accurately detecting surface quality on line in production process of band steel |
CN103399016A (en) * | 2013-07-26 | 2013-11-20 | 齐鲁工业大学 | Online detection system for surface defects of coldly-rolled aluminum plate and detection method of online detection system |
CN105118044A (en) * | 2015-06-16 | 2015-12-02 | 华南理工大学 | Method for automatically detecting defects of wheel-shaped cast product |
CN107316300A (en) * | 2017-07-17 | 2017-11-03 | 杭州盈格信息技术有限公司 | A kind of tire X-ray defect detection method based on depth convolutional neural networks |
CN107403173A (en) * | 2017-08-21 | 2017-11-28 | 合肥麟图信息科技有限公司 | A kind of face identification system and method |
-
2018
- 2018-01-09 CN CN201810020247.8A patent/CN108564104A/en active Pending
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20030202703A1 (en) * | 2002-04-25 | 2003-10-30 | Dainippon Screen Mfg. Co., Ltd. | Apparatus and computer-readable medium for assisting image classification |
CN101996405A (en) * | 2010-08-30 | 2011-03-30 | 中国科学院计算技术研究所 | Method and device for rapidly detecting and classifying defects of glass image |
CN102253049A (en) * | 2011-06-30 | 2011-11-23 | 东北大学 | Method for accurately detecting surface quality on line in production process of band steel |
CN103399016A (en) * | 2013-07-26 | 2013-11-20 | 齐鲁工业大学 | Online detection system for surface defects of coldly-rolled aluminum plate and detection method of online detection system |
CN105118044A (en) * | 2015-06-16 | 2015-12-02 | 华南理工大学 | Method for automatically detecting defects of wheel-shaped cast product |
CN107316300A (en) * | 2017-07-17 | 2017-11-03 | 杭州盈格信息技术有限公司 | A kind of tire X-ray defect detection method based on depth convolutional neural networks |
CN107403173A (en) * | 2017-08-21 | 2017-11-28 | 合肥麟图信息科技有限公司 | A kind of face identification system and method |
Cited By (22)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109509172A (en) * | 2018-09-25 | 2019-03-22 | 无锡动视宫原科技有限公司 | A kind of liquid crystal display flaw detection method and system based on deep learning |
CN109472769A (en) * | 2018-09-26 | 2019-03-15 | 成都数之联科技有限公司 | A kind of bad image defect detection method and system |
CN111060512A (en) * | 2018-10-16 | 2020-04-24 | 希捷科技有限公司 | Intelligent defect identification system |
CN109544522A (en) * | 2018-11-12 | 2019-03-29 | 北京科技大学 | A kind of Surface Defects in Steel Plate detection method and system |
CN109470708A (en) * | 2018-11-30 | 2019-03-15 | 北京百度网讯科技有限公司 | Quality determining method, device, server and the storage medium of plastic foam cutlery box |
CN109584235A (en) * | 2018-11-30 | 2019-04-05 | 北京百度网讯科技有限公司 | Quality determining method, device, server and the storage medium of plastic foam cutlery box |
CN109598712A (en) * | 2018-11-30 | 2019-04-09 | 北京百度网讯科技有限公司 | Quality determining method, device, server and the storage medium of plastic foam cutlery box |
CN109961433A (en) * | 2019-03-29 | 2019-07-02 | 北京百度网讯科技有限公司 | Product defects detection method, device and computer equipment |
CN110554047B (en) * | 2019-09-06 | 2021-07-02 | 腾讯科技(深圳)有限公司 | Method, device, system and equipment for processing product defect detection data |
CN110554047A (en) * | 2019-09-06 | 2019-12-10 | 腾讯科技(深圳)有限公司 | method, device, system and equipment for processing product defect detection data |
CN110738237A (en) * | 2019-09-16 | 2020-01-31 | 深圳新视智科技术有限公司 | Defect classification method and device, computer equipment and storage medium |
CN113632099A (en) * | 2019-12-20 | 2021-11-09 | 京东方科技集团股份有限公司 | Distributed product defect analysis system, method and computer readable storage medium |
WO2021120186A1 (en) * | 2019-12-20 | 2021-06-24 | 京东方科技集团股份有限公司 | Distributed product defect analysis system and method, and computer-readable storage medium |
US11880968B2 (en) | 2019-12-20 | 2024-01-23 | Boe Technology Group Co., Ltd. | Distributed computing system for product defect analysis |
US11719727B1 (en) | 2020-02-10 | 2023-08-08 | National Technology & Engineering Solutions Of Sandia, Llc | Systems and methods for screening particle source manufacturing and development test data |
CN111855668A (en) * | 2020-07-17 | 2020-10-30 | 上海洪朴信息科技有限公司 | Product defect detecting system |
CN112070246A (en) * | 2020-08-26 | 2020-12-11 | 上海洪朴信息科技有限公司 | Photovoltaic product defect detection algorithm deployment system |
CN112464495A (en) * | 2020-12-14 | 2021-03-09 | 广东省科学院智能制造研究所 | Defect detection method and device for mold structure |
CN113129284A (en) * | 2021-02-26 | 2021-07-16 | 山东大学 | Appearance detection method based on 5G cloud edge cooperation and implementation system |
CN112927170A (en) * | 2021-04-08 | 2021-06-08 | 上海哥瑞利软件股份有限公司 | Automatic defect removal method in semiconductor manufacturing process |
CN112927170B (en) * | 2021-04-08 | 2024-03-15 | 上海哥瑞利软件股份有限公司 | Automatic defect removing method in semiconductor manufacturing process |
CN113593142A (en) * | 2021-07-26 | 2021-11-02 | 中国工商银行股份有限公司 | Automatic Teller Machine (ATM) patrolling method and device |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108564104A (en) | Product defects detection method, device, system, server and storage medium | |
KR102229594B1 (en) | Display screen quality detection method, device, electronic device and storage medium | |
KR102321765B1 (en) | Display device quality inspection method, device, electronic device and storage medium | |
CN108230318B (en) | Ladle defects detection classification method, device, equipment and computer-readable medium | |
CN108230317A (en) | Steel plate defect detection sorting technique, device, equipment and computer-readable medium | |
CN108320278A (en) | Product defects detect localization method, device, equipment and computer-readable medium | |
CN108257121B (en) | Method, apparatus, storage medium and the terminal device that product defects detection model updates | |
CN108154508A (en) | Method, apparatus, storage medium and the terminal device of product defects detection positioning | |
CN109084955A (en) | Display screen quality determining method, device, electronic equipment and storage medium | |
CN108830285A (en) | A kind of object detection method of the reinforcement study based on Faster-RCNN | |
CN109085174A (en) | Display screen peripheral circuit detection method, device, electronic equipment and storage medium | |
CN109949286A (en) | Method and apparatus for output information | |
JP2019095217A (en) | Visual inspection device | |
JP2015087903A (en) | Apparatus and method for information processing | |
CN109064446A (en) | Display screen quality determining method, device, electronic equipment and storage medium | |
CN109544548A (en) | Defect inspection method, device, server, equipment and the storage medium of cutlery box | |
CN108257122A (en) | Paper sheet defect detection method, device and server based on machine vision | |
CN112613569B (en) | Image recognition method, training method and device for image classification model | |
CN110264444A (en) | Damage detecting method and device based on weak segmentation | |
CN109087281A (en) | Display screen peripheral circuit detection method, device, electronic equipment and storage medium | |
CN110275820B (en) | Page compatibility testing method, system and equipment | |
CN113393442A (en) | Method and system for detecting abnormality of train parts, electronic device and storage medium | |
KR102174424B1 (en) | Method for Inspecting Component basesd Server and system and apparatus therefor | |
CN113205511A (en) | Electronic component batch information detection method and system based on deep neural network | |
JP2019158684A (en) | Inspection system, identification system, and discriminator evaluation device |
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: 20180921 |
|
RJ01 | Rejection of invention patent application after publication |