CN112964732A - Spinning cake defect visual detection system and method based on deep learning - Google Patents
Spinning cake defect visual detection system and method based on deep learning Download PDFInfo
- Publication number
- CN112964732A CN112964732A CN202110155718.8A CN202110155718A CN112964732A CN 112964732 A CN112964732 A CN 112964732A CN 202110155718 A CN202110155718 A CN 202110155718A CN 112964732 A CN112964732 A CN 112964732A
- Authority
- CN
- China
- Prior art keywords
- spinning cake
- image
- detection
- defect detection
- deep learning
- 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
Images
Classifications
-
- 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
- G01N21/95—Investigating the presence of flaws or contamination characterised by the material or shape of the object to be examined
-
- 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
- G01N21/8806—Specially adapted optical and illumination features
-
- 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
-
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20021—Dividing image into blocks, subimages or windows
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Pathology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Biochemistry (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Immunology (AREA)
- Analytical Chemistry (AREA)
- Chemical & Material Sciences (AREA)
- Artificial Intelligence (AREA)
- Software Systems (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Signal Processing (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Quality & Reliability (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
The invention discloses a spinning cake defect visual detection system and method based on deep learning, comprising a detection platform; the image acquisition module is arranged on the detection table and comprises an end array camera arranged on the end face of the detection table, a side array camera arranged on the side face of the detection table and an array light source assembly arranged on the detection table; and the model training module is used for receiving the image signals to perform model training and is connected to the output end of the image acquisition module. The detection method comprises the steps of collecting a spinning cake image through an image collection module; training a defect detection model by using a model training module; performing spinning cake defect detection on the obtained spinning cake image by using a defect detection model to obtain a spinning cake defect detection result; and (4) summarizing the defects of the spinning cake defect detection results, screening and grading the defects, and outputting a final result. The image acquisition process is simple and efficient, the later maintenance is convenient, the linear array camera can improve the image resolution, the image acquisition quality is high, and the cost is saved.
Description
Technical Field
The invention relates to the technical field of spinning cake defect detection, in particular to a spinning cake defect visual detection system and method based on deep learning.
Background
Spinning cake defect detection is an important link in the spinning cake production process, the spinning cake defect detection is finished in the manual quality detection mode (a flashlight is held by hands, the spinning is polished in multiple directions and multiple angles, and the spinning defect is observed) in the spinning cake industry in China at present, and the manual quality detection mode has many defects, such as high working strength, high possibility of being influenced by subjective factors of people, low efficiency, and difficulty in observing the defect which is slight and unobvious, and the manual quality detection cannot be observed by naked eyes.
The defects of the prior art are that the manual quality inspection is high in cost, high in working strength, large in subjective influence factor, low in efficiency, unobvious in subtle defects, incapable of being observed by naked eyes and the like.
Disclosure of Invention
The invention aims to overcome the defects of the prior art, and in order to realize the purpose, the spinning cake defect visual detection system and method based on deep learning are adopted to solve the problems in the background technology.
A spinning cake defect visual inspection system based on deep learning comprises:
a detection table;
the image acquisition module is arranged on the detection table and comprises an end array camera arranged on the end face of the detection table, a side array camera arranged on the side face of the detection table and an array light source assembly arranged on the detection table;
and the model training module is used for receiving the image signal to perform model training and is connected to the output end of the image acquisition module.
As a further aspect of the invention: the detection table is provided with a manipulator for clamping and fixing.
As a further aspect of the invention: the array light source assembly comprises a right array light source arranged on the detection table and a left array light source arranged on the detection table.
As a further aspect of the invention: and the end real camera and the side array camera both adopt linear array cameras.
A method comprising a deep learning based visual inspection system of spinning cake defects as described in any of the above, comprising the steps of:
collecting a spinning cake image through an image collecting module;
training a defect detection model by using a model training module;
performing spinning cake defect detection on the obtained spinning cake image by using a defect detection model to obtain a spinning cake defect detection result;
and (4) summarizing the defects of the spinning cake defect detection results, screening and grading the defects, and outputting a final result.
As a further aspect of the invention: the specific steps of collecting the spinning cake image through the image collecting module comprise:
the spinning cake to be detected reaches the detection platform, and the mechanical arm moves downwards and clamps the spinning cake and sets the spinning cake to the position to be detected;
and setting the angle of the manipulator and the constant-speed rotating speed of the manipulator, and starting to acquire a plurality of spinning cake images at different angles by the image acquisition module.
As a further aspect of the invention: the specific steps of detecting the spinning cake defects of the spinning cake image comprise:
obtaining a plurality of spinning cake images, and extracting the effective area of each spinning cake image;
dividing the effective area of each spinning cake image into m x n sub-images;
and respectively processing the sub-graphs separated into m × n by using a spinning cake defect detection model, and summarizing the defects.
As a further aspect of the invention: the specific steps of the spinning cake image data preprocessing of the defect detection model comprise:
acquiring spinning cake image data, and initializing a network weight of a defect detection model;
propagating the correlated input data through the full link layer, the downsample layer, and the convolutional layer to form correlated output values;
inputting the characteristics of the spinning cake image data in the convolution kernel to perform characteristic factor processing;
calculating the error between the target value and the output value;
if the error is larger than the expected value, returning back to the network to calculate the error value between layers;
and if the error is less than or equal to the expected value, finishing the training.
As a further aspect of the invention: the specific formula for inputting the characteristics of the spinning cake image data in the convolution kernel to perform characteristic factor processing is as follows:
wherein R is1And Ri+1Input and output, L, representing convolution of layer 1 and layer i +1i+1Is Ri+1R (i, j) represents a pixel of the feature map, f and S0And p are convolutional layer parameters, expressed as convolutional kernel size, convolutional step size, number of filling layers, respectively.
Compared with the prior art, the invention has the following technical effects:
by adopting the technical scheme, the multi-angle linear array camera is used for shooting to obtain image samples with high resolution and good image quality, and meanwhile, the trained defect detection model is used for detecting defects and further defect summarization is carried out. Personnel allocation on spinning cake defect detection is reduced, and labor cost is reduced. The design overcomes the defects of manual quality inspection, and the automatic detection of the spinning cake defects is completed by building a spinning cake image acquisition device and applying a machine vision technology and a deep learning algorithm.
Drawings
The following detailed description of embodiments of the invention refers to the accompanying drawings in which:
FIG. 1 is a schematic view of a visual inspection system for spinning cake defects according to some embodiments disclosed herein;
FIG. 2 is a block flow diagram of a visual inspection method of spinning cake defects according to some embodiments disclosed herein;
FIG. 3 is a schematic diagram of a connection between an image acquisition module and a model training module according to some embodiments disclosed herein.
In the figure: 1. a detection table; 2. an image acquisition module; 21. an end array camera; 22. a side array camera; 23. an array light source assembly; 231. a right array light source; 232. a left array light source; 3. and a model training module.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, in an embodiment of the present invention, a visual inspection system for spinning cake defects based on deep learning includes:
a detection table 1;
set up in examine test table 1 and be used for gathering the image acquisition module 2 of each angle image of spinning cake, image acquisition module 2 is including setting up in examining the 1 terminal surface of test table and being used for gathering the tip array camera 21 of upper and lower surface image, tip array camera 21 is provided with up and down the terminal surface and respectively has a tip array camera 21, sets up in examining 1 side of test table and is used for spinning cake side and oblique side image acquisition's lateral part array camera 22, lateral part array camera 22 is equipped with 1 to 3 lateral part array camera 22, just tip real camera and lateral part array camera 22 all adopt the linear array camera. And an array light source assembly 23 arranged on the detection table 1 and used for supplementing the brightness of the ambient light source; the array light source assembly 23 comprises a right array light source 231 arranged on the detection table 1 and a left array light source 232 arranged on the detection table 1.
As shown in fig. 3, the model training module 3 is configured to receive an image signal for model training, and the model training module 3 is connected to an output end of the image acquisition module 2.
In some specific embodiments, the inspection station 1 is provided with a robot for gripping and holding a sample of a spinning cake.
In this embodiment, 5 line-scan cameras can be selected to collect 5 images, wherein 2 line-scan cameras on the inclined side are responsible for detecting defects of paper tubes, 2 line-scan cameras on the upper and lower end faces are responsible for detecting defects of spinning cakes on the upper and lower surfaces, and 1 line-scan camera on the side face is responsible for detecting defects of the spinning cake side face.
The working principle and the working process of some embodiments disclosed by the invention are as follows:
the spinning cake to be detected reaches a detection platform 1;
the clamping jaw of the manipulator is downward and clamped after being in place;
after clamping, the clamping jaw rises to the highest position, after the clamping jaw is static for 1S, the clamping jaw starts to rotate at a constant speed according to a set angle and speed, at the moment, a PLC sends a pattern acquisition signal to an upper computer vision detection software system, and 5 linear array cameras start to acquire patterns;
after the set rotation angle is reached, the image acquisition is finished, the upper computer vision detection software system carries out defect detection on the acquired image, meanwhile, the clamping jaw is downwards placed on the spinning cake, and after the spinning cake reaches a low position, the clamping jaw is loosened and ascended;
after the defect detection is finished, the information is stored in a local database, and the spinning cake defect information is automatically transmitted to an ERP (enterprise resource planning), WMS (warehouse management system), TMS (intelligent distribution management system) and other systems, so that one-time detection is finished.
In some specific embodiments, the image acquisition module 2 of the inspection station 1 may be split into 2 parts.
One is as follows: the collecting station 1 comprises 3 linear array cameras and 1 linear array light source, 2 linear array cameras are used for detecting the defects of the upper surface and the lower surface of the spinning cake and are used for detecting the defects of the side surface of the spinning cake, and 1 linear array camera on the side surface is used for detecting the defects of the side surface of the spinning cake.
The second step is as follows: and (3) a collection station 2: 4 linear array cameras, 2 array light sources, 4 linear array cameras are responsible for detecting the paper tube defect.
As shown in fig. 2, a method comprising a deep learning based visual inspection system for spinning cake defects as described above, comprising the steps of:
s1, acquiring a spinning cake image through the image acquisition module 2;
when the spinning cake to be detected reaches the detection table 1, the manipulator moves downwards and clamps the spinning cake, and then the spinning cake is set to the position to be detected;
the angle of the manipulator and the uniform rotation speed thereof are set, and the image acquisition module 2 starts to acquire a plurality of spinning cake images at different angles.
S2, training a defect detection model by using the model training module 3;
s3, performing spinning cake defect detection on the obtained spinning cake image by using a defect detection model to obtain a spinning cake defect detection result;
obtaining a plurality of spinning cake images, and extracting the effective area of each spinning cake image;
dividing the effective area of each spinning cake image into m x n sub-images;
and respectively processing the sub-graphs separated into m × n by using a spinning cake defect detection model, and summarizing the defects.
And S4, collecting the defects of the spinning cake defect detection results, screening and grading the defects, and outputting the final result.
In some specific embodiments, the defect detection models include a paper tube defect detection model, a surface defect detection model, and a side defect detection model.
In some specific embodiments, the specific steps of the cake image data preprocessing of the defect detection model include:
acquiring spinning cake image data, and initializing a network weight of a defect detection model;
propagating the correlated input data through the full link layer, the downsample layer, and the convolutional layer to form correlated output values;
inputting the characteristics of the spinning cake image data in the convolution kernel to perform characteristic factor processing;
calculating the error between the target value and the output value;
if the error is larger than the expected value, returning back to the network to calculate the error value between layers;
and if the error is less than or equal to the expected value, finishing the training.
The specific formula for inputting the characteristics of the spinning cake image data in the convolution kernel to perform characteristic factor processing is as follows:
wherein R is1And Ri+1Input and output, L, representing convolution of layer 1 and layer i +1i+1Is Ri+1R (i, j) represents a pixel of the feature map, f and S0And p are convolutional layer parameters, expressed as convolutional kernel size, convolutional step size, number of filling layers, respectively.
In this embodiment, the specific steps of cake image data preprocessing based on deep learning include:
turning the collected spinning cake image data, performing secondary sampling to generate 2D image data based on the original drawing, and positioning corresponding positions of specific defects in the data through classification based on a deep learning network.
Initializing related network weight;
propagating the correlated input data through the full link layer, the downsample layer, and the convolutional layer to form correlated output values;
calculating the error between the target value and the output value;
if the error is higher than the expected value, the error is transmitted back to the network, and the error values among the three types of layers are calculated in sequence. If the error does not exceed the expected value, the training is finished.
Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents, which should be construed as being within the scope of the invention.
Claims (9)
1. A spinning cake defect visual inspection system based on deep learning is characterized by comprising:
a detection table (1);
the image acquisition module (2) is arranged on the detection table and comprises an end array camera (21) arranged on the end face of the detection table, a side array camera (22) arranged on the side face of the detection table and an array light source assembly (23) arranged on the detection table;
and the model training module (3) is used for receiving the image signal to perform model training and is connected to the output end of the image acquisition module.
2. The deep learning-based visual inspection system for spinning cake defects according to claim 1, wherein the inspection table is provided with a manipulator for clamping and fixing.
3. The deep learning-based visual spinning cake defect detection system as claimed in claim 2, wherein the array light source assembly comprises a right array light source (231) arranged on the detection table and a left array light source (232) arranged on the detection table.
4. The deep learning-based visual inspection system for spinning cake defects according to claim 1, wherein both the end real camera and the side array camera employ line cameras.
5. A method comprising a deep learning based visual inspection system of spinning cake defects according to any of claims 1 to 4, comprising the steps of:
collecting a spinning cake image through an image collecting module;
training a defect detection model by using a model training module;
performing spinning cake defect detection on the obtained spinning cake image by using a defect detection model to obtain a spinning cake defect detection result;
and (4) summarizing the defects of the spinning cake defect detection results, screening and grading the defects, and outputting a final result.
6. The visual inspection system for spinning cake defects based on deep learning as claimed in claim 5, wherein the specific steps of collecting spinning cake images through the image collection module comprise:
the spinning cake to be detected reaches the detection platform, and the mechanical arm moves downwards and clamps the spinning cake and sets the spinning cake to the position to be detected;
and setting the angle of the manipulator and the constant-speed rotating speed of the manipulator, and starting to acquire a plurality of spinning cake images at different angles by the image acquisition module.
7. The deep learning-based visual inspection system for spinning cake defects according to claim 6, wherein the specific steps of performing spinning cake defect inspection on the spinning cake images comprise:
obtaining a plurality of spinning cake images, and extracting the effective area of each spinning cake image;
dividing the effective area of each spinning cake image into m x n sub-images;
and respectively processing the sub-graphs separated into m × n by using a spinning cake defect detection model, and summarizing the defects.
8. The deep learning-based visual spinning cake defect detection system as claimed in claim 7, wherein the specific steps of the pretreatment of the spinning cake image data of the defect detection model comprise:
acquiring spinning cake image data, and initializing a network weight of a defect detection model;
propagating the correlated input data through the full link layer, the downsample layer, and the convolutional layer to form correlated output values;
inputting the characteristics of the spinning cake image data in the convolution kernel to perform characteristic factor processing;
calculating the error between the target value and the output value;
if the error is larger than the expected value, returning back to the network to calculate the error value between layers;
and if the error is less than or equal to the expected value, finishing the training.
9. The visual inspection system for spinning cake defects based on deep learning as claimed in claim 8, wherein the specific formula for feature element processing of the features of the input spinning cake image data in the convolution kernel is as follows:
wherein R is1And Ri+1Input and output, L, representing convolution of layer 1 and layer i +1i+1Is Ri+1R (i, j) represents a pixel of the feature map, f and S0And p are convolutional layer parameters, expressed as convolutional kernel size, convolutional step size, number of filling layers, respectively.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110155718.8A CN112964732A (en) | 2021-02-04 | 2021-02-04 | Spinning cake defect visual detection system and method based on deep learning |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110155718.8A CN112964732A (en) | 2021-02-04 | 2021-02-04 | Spinning cake defect visual detection system and method based on deep learning |
Publications (1)
Publication Number | Publication Date |
---|---|
CN112964732A true CN112964732A (en) | 2021-06-15 |
Family
ID=76273870
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110155718.8A Pending CN112964732A (en) | 2021-02-04 | 2021-02-04 | Spinning cake defect visual detection system and method based on deep learning |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112964732A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114252452A (en) * | 2021-12-22 | 2022-03-29 | 西安交通大学 | Online detection device and method for appearance defects and contour dimension of small-sized revolving body |
CN116167992A (en) * | 2023-02-02 | 2023-05-26 | 北京百度网讯科技有限公司 | Defect detection method, defect detection device, electronic device and computer readable storage medium |
WO2023168972A1 (en) * | 2022-03-09 | 2023-09-14 | 三门三友科技股份有限公司 | Linear array camera-based copper surface defect detection method and apparatus |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108362701A (en) * | 2018-05-18 | 2018-08-03 | 西安获德图像技术有限公司 | A kind of chemical fiber spinning cake appearance automatic checkout system |
CN109454006A (en) * | 2018-10-08 | 2019-03-12 | 杭州慧知连科技有限公司 | Chemical fiber wire ingot is stumbled the device and its detection stage division of a defect on-line checking and classification |
CN111028160A (en) * | 2019-11-21 | 2020-04-17 | 西北工业大学 | Remote sensing image noise suppression method based on convolutional neural network |
CN111652852A (en) * | 2020-05-08 | 2020-09-11 | 浙江华睿科技有限公司 | Method, device and equipment for detecting surface defects of product |
CN212301358U (en) * | 2020-04-16 | 2021-01-05 | 杭州慧知连科技有限公司 | Visual linear scanning detection device for surface defects of silk ingots |
-
2021
- 2021-02-04 CN CN202110155718.8A patent/CN112964732A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108362701A (en) * | 2018-05-18 | 2018-08-03 | 西安获德图像技术有限公司 | A kind of chemical fiber spinning cake appearance automatic checkout system |
CN109454006A (en) * | 2018-10-08 | 2019-03-12 | 杭州慧知连科技有限公司 | Chemical fiber wire ingot is stumbled the device and its detection stage division of a defect on-line checking and classification |
CN111028160A (en) * | 2019-11-21 | 2020-04-17 | 西北工业大学 | Remote sensing image noise suppression method based on convolutional neural network |
CN212301358U (en) * | 2020-04-16 | 2021-01-05 | 杭州慧知连科技有限公司 | Visual linear scanning detection device for surface defects of silk ingots |
CN111652852A (en) * | 2020-05-08 | 2020-09-11 | 浙江华睿科技有限公司 | Method, device and equipment for detecting surface defects of product |
Non-Patent Citations (2)
Title |
---|
王泽霞 等: "基于改进卷积神经网络的化纤丝饼表面缺陷识别", 《纺织学报》 * |
秦川: "基于卷积神经网络的图像识别", 《电子技术与软件工程》 * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114252452A (en) * | 2021-12-22 | 2022-03-29 | 西安交通大学 | Online detection device and method for appearance defects and contour dimension of small-sized revolving body |
WO2023168972A1 (en) * | 2022-03-09 | 2023-09-14 | 三门三友科技股份有限公司 | Linear array camera-based copper surface defect detection method and apparatus |
CN116167992A (en) * | 2023-02-02 | 2023-05-26 | 北京百度网讯科技有限公司 | Defect detection method, defect detection device, electronic device and computer readable storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN112964732A (en) | Spinning cake defect visual detection system and method based on deep learning | |
CN109724984B (en) | Defect detection and identification device and method based on deep learning algorithm | |
CN108090896B (en) | Wood board flatness detection and machine learning method and device and electronic equipment | |
CN107643295A (en) | A kind of method and system of the cloth defect on-line checking based on machine vision | |
CN104483320B (en) | Digitized defect detection device and detection method of industrial denitration catalyst | |
CN109840900A (en) | A kind of line detection system for failure and detection method applied to intelligence manufacture workshop | |
CN102175692A (en) | System and method for detecting defects of fabric gray cloth quickly | |
CN114638797A (en) | Method and device for detecting copper surface defects based on linear array camera | |
CN110186375A (en) | Intelligent high-speed rail white body assemble welding feature detection device and detection method | |
CN110020691A (en) | LCD screen defect inspection method based on the training of convolutional neural networks confrontation type | |
CN110118777A (en) | A kind of control system system integration Smart Verify platform | |
CN114029237A (en) | Real-time sorting system and method for surface defects of aluminum profiles | |
CN110503623A (en) | A method of Bird's Nest defect on the identification transmission line of electricity based on convolutional neural networks | |
CN112819780A (en) | Method and system for detecting surface defects of silk ingots and silk ingot grading system | |
CN111504381A (en) | Terminal row product online detection device and detection method thereof | |
CN112730425A (en) | Industrial scene oriented product surface defect automatic identification method | |
CN113588653A (en) | System and method for detecting and tracking quality of aluminum anode carbon block | |
CN114529510B (en) | Automatic detection and classification method for cathode copper on-line quality | |
CN110728269B (en) | High-speed rail contact net support pole number plate identification method based on C2 detection data | |
CN113888408A (en) | Multi-camera image acquisition method and device | |
CN113870204A (en) | Method and device for detecting abnormality of optical glass | |
CN116091506A (en) | Machine vision defect quality inspection method based on YOLOV5 | |
CN115060742A (en) | Printed circuit board defect detection system and method based on visual edge calculation | |
CN211043180U (en) | Optical cable detection machine and optical cable manufacturing machine | |
CN110070520B (en) | Pavement crack detection model construction and detection method based on deep neural network |
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 |