CN111768380A - Method for detecting surface defects of industrial spare and accessory parts - Google Patents
Method for detecting surface defects of industrial spare and accessory parts Download PDFInfo
- Publication number
- CN111768380A CN111768380A CN202010605154.9A CN202010605154A CN111768380A CN 111768380 A CN111768380 A CN 111768380A CN 202010605154 A CN202010605154 A CN 202010605154A CN 111768380 A CN111768380 A CN 111768380A
- Authority
- CN
- China
- Prior art keywords
- network
- data
- surface defects
- sample
- steps
- 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
- 230000007547 defect Effects 0.000 title claims abstract description 63
- 238000000034 method Methods 0.000 title claims abstract description 62
- 238000001514 detection method Methods 0.000 claims abstract description 44
- 238000012549 training Methods 0.000 claims abstract description 33
- 230000006870 function Effects 0.000 claims description 17
- 238000000605 extraction Methods 0.000 claims description 14
- 238000003860 storage Methods 0.000 claims description 7
- 238000010606 normalization Methods 0.000 claims description 4
- 238000007781 pre-processing Methods 0.000 claims description 4
- 238000002360 preparation method Methods 0.000 claims description 4
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 claims description 3
- 241000272525 Anas platyrhynchos Species 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 3
- 238000005520 cutting process Methods 0.000 claims description 2
- 238000007689 inspection Methods 0.000 claims 2
- 238000013473 artificial intelligence Methods 0.000 abstract description 2
- 238000004519 manufacturing process Methods 0.000 description 8
- 238000002372 labelling Methods 0.000 description 4
- 230000002950 deficient Effects 0.000 description 3
- 206010063385 Intellectualisation Diseases 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 238000004590 computer program Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- 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
- G06N20/00—Machine learning
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Quality & Reliability (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Medical Informatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention relates to the technical field of information data artificial intelligence, in particular to a method for detecting surface defects of industrial spare and accessory parts, which comprises the following steps: acquiring images of a normal sample and a defect sample, and marking the samples; expanding training sample data according to the marked sample, generating expanded sample data and normalizing image data; constructing a framework of a collaborative learning defect detection model, and performing model training by using extension sample data; carrying out defect detection on the new image by using the trained model; compared with the prior art, the method can still detect the surface defects of the industrial parts under the condition of marking a small amount of samples, has strong generalization capability and high detection accuracy, is practical and convenient, and not only can be used for detecting the surface defects of the industrial parts, but also can be used for detecting the surface conditions of any other parts/objects.
Description
Technical Field
The invention relates to the technical field of information data artificial intelligence, in particular to a method for detecting surface defects of industrial spare and accessory parts.
Background
Along with the development of the manufacturing industry, the intellectualization and automation of manufacturing and detection become necessary requirements for promoting the high-quality development of the manufacturing industry, and the surface defect detection of industrial parts is an important link for replacing manual visual detection, improving the efficiency, reducing the production cost and realizing the intellectualization and the automation in the production line of the manufacturing industry, however, the traditional automatic optical detection method adopts a manual characteristic extraction mode, is greatly limited by a specific environment and has poor universality and generalization performance. The existing defect detection methods based on deep learning need a large amount of labeled data in the model construction and iteration processes, and the problems of high manual labeling cost and few available samples at the early stage in the actual production process also limit the application of the methods.
Therefore, it is necessary to provide a new method for detecting surface component defects with high accuracy by using a small amount of marked samples.
Disclosure of Invention
The invention breaks through the difficult problems in the prior art, designs the industrial spare and accessory part surface defect detection method based on collaborative learning, and aims to obtain a model with strong generalization capability and high accuracy on the premise of marking a small amount of samples, thereby replacing the artificial visual defect detection in a production line.
In order to achieve the purpose, the invention designs a method for detecting surface defects of industrial parts, which is characterized by comprising the following steps: the detection is carried out according to the following steps:
s1, acquiring images of a normal sample and a defect sample, and marking the samples;
s2, according to the marked sample, expanding the training sample data, generating expanded sample data and normalizing the image data;
s3, constructing a framework of a collaborative learning defect detection model, and performing model training by using the extended sample data;
s4 performs defect detection using the trained model for the new image.
Further, the expansion method of the training sample data in S2 includes one or more of affine transformation, random cropping and duck forced-feeding.
Further, the collaborative learning defect detection model in S3 includes a feature extraction network, a primary network, and a secondary network.
Further, the training method of the collaborative learning defect detection model in S3 is as follows: firstly, inputting the characteristics extracted by an image through a characteristic extraction network into a main network and a secondary network respectively for sharing and using; and then the main network outputs an object recognition result, and the auxiliary network simultaneously outputs an image classification result and an object recognition result.
When the input image has no object-level label, the object recognition output result of the auxiliary network is used as a pseudo label of the main network to assist the main network to complete the process of supervised learning;
when the input image has object-level labels, the object recognition output result of the main network can know the learning process of the auxiliary network. And enabling the main network and the auxiliary network to perform collaborative learning through the loss function.
Further, the loss functions include image classification loss, object recognition loss, and consistency loss.
Further, the function of the consistency loss constraint is defined as:
wherein Bw and Bs respectively represent the set of target frames detected by the primary network and the secondary network in a batch, i represents the ith target frame in Bs, and j represents the jth target frame in Bw. p is a radical ofjAnd piFor output corresponding to the category to which the target box belongs, tjAnd tiFor the corresponding location prediction result, w represents the primary network, s represents the secondary network β is a hyper-parameter, whose value is (0, 1). R represents the regularization function:
the invention also designs a detection system for the surface defects of the industrial spare and accessory parts, which is characterized in that: the defect detection system comprises a data preparation unit for acquiring original data and marking the data, a data preprocessing unit for enhancing and normalizing the data, and a model unit for constructing and training a defect detection model for collaborative learning.
The invention also designs computer equipment which is characterized in that: comprises a processor; a memory for storing processor-executable instructions; wherein the processor is configured to perform the following method steps: acquiring images of a normal sample and a defect sample, and marking the samples; expanding training sample data according to the marked sample, generating expanded sample data and normalizing the image data; constructing a framework of a collaborative learning defect detection model, and performing model training by using extension sample data; and carrying out defect detection by using the trained model aiming at the new image.
Further, the training method of the collaborative learning defect detection model comprises the following steps: firstly, inputting the characteristics extracted by an input image through a characteristic extraction network into a main network and an auxiliary network respectively for sharing and using; and then the main network outputs an object recognition result, and the auxiliary network simultaneously outputs an image classification result and an object recognition result.
When the input image has no object-level label, the object recognition output result of the auxiliary network is used as a pseudo label of the main network to assist the main network to complete the process of supervised learning;
when the input image has object-level labels, the object recognition output result of the main network can know the learning process of the auxiliary network. And enabling the main network and the auxiliary network to perform collaborative learning through the loss function.
A computer-readable storage medium characterized by: computer-executable instructions are stored within the computer-readable storage medium, which when executed by a processor, cause the processor to perform the method steps of: acquiring images of a normal sample and a defect sample, and marking the samples; expanding training sample data according to the marked sample, generating expanded sample data and normalizing the image data; constructing a framework of a collaborative learning defect detection model, and performing model training by using extension sample data; and carrying out defect detection on the new image by using the trained model.
Further, the training method of the collaborative learning defect detection model comprises the following steps: firstly, inputting the characteristics extracted by an input image through a characteristic extraction network into a main network and an auxiliary network respectively for sharing and using; and then the main network outputs an object recognition result, and the auxiliary network simultaneously outputs an image classification result and an object recognition result.
When the input image has no object-level label, the object recognition output result of the auxiliary network is used as a pseudo label of the main network to assist the main network to complete the process of supervised learning;
when the input image has object-level labels, the object recognition output result of the main network can know the learning process of the auxiliary network. And enabling the main network and the auxiliary network to perform collaborative learning through the loss function.
Compared with the prior art, the method can still detect the surface defects of the industrial parts under the condition of marking a small amount of samples, has strong generalization capability and high detection accuracy, is practical and convenient, and can be used for detecting the surface defects of the industrial parts and the surface conditions of any other parts/objects.
Drawings
Fig. 1 is a schematic flow chart of a method for detecting surface defects of an industrial component according to an embodiment of the present invention.
Fig. 2 is a schematic flow chart illustrating training of the collaborative learning defect detection model according to an embodiment of the present invention.
Fig. 3 is a schematic structural diagram of a system for detecting surface defects of an industrial component according to an embodiment of the present invention.
Wherein, 1 is a data preparation unit, 2 is a data preprocessing unit, and 3 is a model unit.
Detailed Description
The invention will be further described with reference to the accompanying drawings, but is not to be construed as being limited thereto.
Referring to fig. 1, in an embodiment of the present invention, M1 normal sample pictures and M2 defective sample pictures are taken and subjected to picture-level labeling, i.e., whether the pictures are normal or defective.
Then, M3 defect samples are selected from M2 defect sample pictures to mark the object level, that is, mark the specific position of the defect, it should be noted that the number of M3 should not be greater than M2, and may even be much smaller than M2.
Then, image data enhancement is carried out on all pictures through affine transformation, random cutting, duck force filling method and other methods, an original data set is expanded, image normalization processing is carried out on the expanded original data set, and finally a training sample set { N } is obtainediAnd (i ∈ 1,2 … k1 … k2 … m), wherein m represents the total number of pictures, k1 represents the number of normal pictures, k2 represents the number of pictures which are not subjected to object-level labeling, namely, m-k1 defective pictures, and the number of pictures in which object-level labeling is performed is m-k1-k 2.
Then starting a component collaborative learning defect detection model which comprises 3 network components of a feature extraction network, a main network and a secondary network, wherein the feature extraction network can use the feature extraction network commonly used in object recognition, and then inputting a training sample set { N }iAnd fifthly, carrying out model training.
Referring to FIG. 2, the input image is the training sample set { N }iAnd inputting the characteristics of the input image extracted by the characteristic extraction network into the main network and the auxiliary network respectively for sharing use.
The main network outputs an object recognition result, and the auxiliary network simultaneously outputs an image classification result and an object recognition result. And when the input image has no object-level label, the object identification output of the auxiliary network is used as a pseudo label of the main network, and the auxiliary network assists the main network to complete the process of supervised learning. When the input image has object-level labels, the output result of the main network guides the learning process of the auxiliary network. Both learn synergistically through consistency loss constraints.
The loss function of the model training process comprises 3 loss functions of image classification loss, object recognition loss and consistency loss, wherein the image classification loss and the object recognition loss select corresponding loss functions according to the characteristic extraction network, and the consistency loss function is defined as follows:
wherein Bw and Bs respectively represent the set of target frames detected by the primary network and the secondary network in a batch, i represents the ith target frame in Bs, and j represents the jth target frame in Bw. p is a radical ofjAnd piFor output corresponding to the category to which the target box belongs, tjAnd tiFor the corresponding location prediction result, w represents the primary network, s represents the secondary network β is a hyper-parameter, whose value is (0, 1). R represents the regularization function:
in order to complete the detection method, the invention also designs a detection system for the surface defects of the industrial parts, which comprises a data preparation unit 1 for acquiring original data and carrying out data marking, a data preprocessing unit 2 for carrying out data enhancement and normalization, and a model unit 3 for constructing and training a defect detection model in cooperation with learning.
It should be noted that: and in the model training process, the main network and the auxiliary network are used for training at the same time, and in the model using process, the auxiliary network does not work and only the output of the main network is taken as a final detection result.
In a specific embodiment, the present invention also contemplates a computer device comprising a processor; a memory for storing processor-executable instructions; wherein the processor is configured to perform the following method steps: acquiring images of a normal sample and a defect sample, and marking the samples; expanding training sample data according to the marked sample, generating expanded sample data and normalizing image data; constructing a framework of a collaborative learning defect detection model, and performing model training by using extension sample data; and carrying out defect detection by using the trained model aiming at the new image.
Preferably, the training method of the collaborative learning defect detection model comprises the following steps: firstly, inputting the characteristics extracted by an input image through a characteristic extraction network into a main network and an auxiliary network respectively for sharing and using; and then the main network outputs an object recognition result, and the auxiliary network simultaneously outputs an image classification result and an object recognition result.
When the input image has no object-level label, the object recognition output result of the auxiliary network is used as a pseudo label of the main network to assist the main network to complete the process of supervised learning;
when the input image has object-level labels, the object recognition output result of the main network can know the learning process of the auxiliary network. And enabling the main network and the auxiliary network to perform collaborative learning through the loss function.
In a specific embodiment, the present invention contemplates a computer-readable storage medium having stored therein computer-executable instructions that, when executed by a processor, cause the processor to perform the method steps of: acquiring images of a normal sample and a defect sample, and marking the samples; according to the marked sample, carrying out extension of training sample data, generating extension sample data and carrying out normalization of image data; constructing a framework of a collaborative learning defect detection model, and performing model training by using extension sample data; and carrying out defect detection by using the trained model aiming at the new image.
Preferably, the training method of the collaborative learning defect detection model comprises the following steps: firstly, inputting the characteristics extracted by an input image through a characteristic extraction network into a main network and an auxiliary network respectively for sharing and using; and then the main network outputs an object recognition result, and the auxiliary network simultaneously outputs an image classification result and an object recognition result.
When the input image has no object-level label, the object recognition output result of the auxiliary network is used as a pseudo label of the main network to assist the main network to complete the process of supervised learning;
when the input image has object-level labels, the object recognition output result of the main network can know the learning process of the auxiliary network. And enabling the main network and the auxiliary network to perform collaborative learning through the loss function.
It will be understood by those skilled in the art that all or part of the processes of the methods of the above embodiments may be implemented by hardware related to computer program instructions, and the program may be stored in a computer readable storage medium, and may be executed by at least one processor of a computer system to implement the processes of the embodiments including the methods of the above embodiments according to the embodiments of the present invention. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), or the like.
The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above-mentioned embodiments only express several embodiments of the present invention, and the description is specific and detailed, but not understood as the scope of the present invention is limited thereby, it should be noted that, for those skilled in the art, many variations and modifications can be made without departing from the concept of the present invention, and these all fall into the protection scope of the present invention.
Claims (9)
1. A method for detecting surface defects of industrial parts is characterized by comprising the following steps: the detection is carried out according to the following steps:
s1, acquiring images of a normal sample and a defect sample, and marking the samples;
s2, according to the marked sample, expanding the training sample data, generating expanded sample data and normalizing the image data;
s3, constructing a framework of a collaborative learning defect detection model, and performing model training by using extension sample data;
s4 performs defect detection using the trained model for the new image.
2. The method for detecting surface defects of industrial parts according to claim 1, wherein the method comprises the following steps: the expansion method of the training sample data in the S2 comprises one or more of affine transformation, random cutting and duck forced-feeding method.
3. The method for detecting surface defects of industrial parts according to claim 1, wherein the method comprises the following steps: the collaborative learning defect detection model in the S3 comprises a feature extraction network, a main network and an auxiliary network.
4. The method for detecting surface defects of industrial parts according to claim 1, wherein the method comprises the following steps: the training method of the collaborative learning defect detection model in the S3 comprises the following steps: firstly, inputting the characteristics extracted by an input image through a characteristic extraction network into a main network and an auxiliary network respectively for sharing and using; then the main network outputs an object recognition result, and the auxiliary network simultaneously outputs an image classification result and an object recognition result;
when the input image has no object-level label, the object recognition output result of the auxiliary network is used as a pseudo label of the main network to assist the main network to complete the process of supervised learning;
when the input image has object-level labels, the learning process of the auxiliary network can be known by the object identification output result of the main network;
and the main network and the auxiliary network carry out collaborative learning through the loss function.
5. The method for detecting surface defects of industrial parts according to claim 4, wherein the method comprises the following steps: the loss functions include image classification loss, object recognition loss, and consistency loss.
6. The method for detecting surface defects of industrial parts according to claim 5, wherein the method comprises the following steps: the function of the consistency loss constraint is defined as:
where Bw and Bs denote primary and secondary network detection, respectively, in one batchI represents the ith target box in Bs, and j represents the jth target box in Bw. p is a radical ofjAnd piFor output corresponding to the category to which the target box belongs, tjAnd tiFor the corresponding location prediction result, w represents the primary network, s represents the secondary network β is a hyper-parameter, whose value is (0, 1). R represents the regularization function:
7. the inspection system for industrial component surface defect inspection method according to any one of claims 1 to 6, wherein: the method comprises a data preparation unit (1) for acquiring original data and performing data marking, a data preprocessing unit (2) for performing data enhancement and normalization, and a model unit (3) for constructing and training a collaborative learning defect detection model.
8. A computer device, characterized by: comprises a processor;
a memory for storing processor-executable instructions;
wherein the processor is configured to: the method of any one of claims 1-6 may be performed.
9. A computer-readable storage medium characterized by: the computer-readable storage medium has stored therein computer-executable instructions that, when executed by a processor, cause the processor to perform the steps of the method of any one of claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010605154.9A CN111768380A (en) | 2020-06-29 | 2020-06-29 | Method for detecting surface defects of industrial spare and accessory parts |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010605154.9A CN111768380A (en) | 2020-06-29 | 2020-06-29 | Method for detecting surface defects of industrial spare and accessory parts |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111768380A true CN111768380A (en) | 2020-10-13 |
Family
ID=72724395
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010605154.9A Pending CN111768380A (en) | 2020-06-29 | 2020-06-29 | Method for detecting surface defects of industrial spare and accessory parts |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111768380A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113160192A (en) * | 2021-04-28 | 2021-07-23 | 北京科技大学 | Visual sense-based snow pressing vehicle appearance defect detection method and device under complex background |
| CN114813730A (en) * | 2022-03-10 | 2022-07-29 | 慧之安信息技术股份有限公司 | Method and device for detecting industrial zero device based on edge calculation |
| CN116563273A (en) * | 2023-06-30 | 2023-08-08 | 张家港广大特材股份有限公司 | Detection and early warning method and system for steel defects |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050234955A1 (en) * | 2004-04-15 | 2005-10-20 | Microsoft Corporation | Clustering based text classification |
| CN108734195A (en) * | 2018-04-13 | 2018-11-02 | 王延峰 | Weakly supervised detection model training method based on Cooperative Study and system |
| CN109636772A (en) * | 2018-10-25 | 2019-04-16 | 同济大学 | The defect inspection method on the irregular shape intermetallic composite coating surface based on deep learning |
| CN109683360A (en) * | 2019-02-14 | 2019-04-26 | 宁波轻蜓视觉科技有限公司 | Liquid crystal display panel defect inspection method and device |
| WO2019136946A1 (en) * | 2018-01-15 | 2019-07-18 | 中山大学 | Deep learning-based weakly supervised salient object detection method and system |
| CN110533086A (en) * | 2019-08-13 | 2019-12-03 | 天津大学 | The semi-automatic mask method of image data |
| CN110570410A (en) * | 2019-09-05 | 2019-12-13 | 河北工业大学 | Detection method for automatically identifying and detecting weld defects |
-
2020
- 2020-06-29 CN CN202010605154.9A patent/CN111768380A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050234955A1 (en) * | 2004-04-15 | 2005-10-20 | Microsoft Corporation | Clustering based text classification |
| WO2019136946A1 (en) * | 2018-01-15 | 2019-07-18 | 中山大学 | Deep learning-based weakly supervised salient object detection method and system |
| CN108734195A (en) * | 2018-04-13 | 2018-11-02 | 王延峰 | Weakly supervised detection model training method based on Cooperative Study and system |
| CN109636772A (en) * | 2018-10-25 | 2019-04-16 | 同济大学 | The defect inspection method on the irregular shape intermetallic composite coating surface based on deep learning |
| CN109683360A (en) * | 2019-02-14 | 2019-04-26 | 宁波轻蜓视觉科技有限公司 | Liquid crystal display panel defect inspection method and device |
| CN110533086A (en) * | 2019-08-13 | 2019-12-03 | 天津大学 | The semi-automatic mask method of image data |
| CN110570410A (en) * | 2019-09-05 | 2019-12-13 | 河北工业大学 | Detection method for automatically identifying and detecting weld defects |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113160192A (en) * | 2021-04-28 | 2021-07-23 | 北京科技大学 | Visual sense-based snow pressing vehicle appearance defect detection method and device under complex background |
| CN114813730A (en) * | 2022-03-10 | 2022-07-29 | 慧之安信息技术股份有限公司 | Method and device for detecting industrial zero device based on edge calculation |
| CN116563273A (en) * | 2023-06-30 | 2023-08-08 | 张家港广大特材股份有限公司 | Detection and early warning method and system for steel defects |
| CN116563273B (en) * | 2023-06-30 | 2023-10-27 | 张家港广大特材股份有限公司 | Detection and early warning method and system for steel defects |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109741332B (en) | Man-machine cooperative image segmentation and annotation method | |
| CN111768380A (en) | Method for detecting surface defects of industrial spare and accessory parts | |
| CN110533086B (en) | Semi-automatic image data labeling method | |
| CN111179251A (en) | Defect detection system and method based on twin neural network and by utilizing template comparison | |
| CN111709456A (en) | Wine bottle surface defect detection method, electronic device and storage medium | |
| CN110705630A (en) | Semi-supervised learning type target detection neural network training method, device and application | |
| CN113344888A (en) | Surface defect detection method and device based on combined model | |
| US20220076404A1 (en) | Defect management apparatus, method and non-transitory computer readable medium | |
| CN113506239B (en) | Strip steel surface defect detection method based on cross-stage local network | |
| CN112861861B (en) | Method and device for recognizing nixie tube text and electronic equipment | |
| CN110599453A (en) | Panel defect detection method and device based on image fusion and equipment terminal | |
| CN116596896A (en) | Detection method and system for weld defects based on small sample target detection network | |
| CN117516937A (en) | Rolling bearing unknown fault detection method based on multi-mode feature fusion enhancement | |
| CN114299040A (en) | Ceramic tile flaw detection method and device and electronic equipment | |
| CN112766351A (en) | Image quality evaluation method, system, computer equipment and storage medium | |
| CN111709451A (en) | Wine bottle surface defect detection method, electronic device and storage medium | |
| CN115984215A (en) | Fiber bundle defect detection method based on twin network | |
| CN115100451B (en) | Data expansion method for monitoring oil leakage of hydraulic pump | |
| CN114529517B (en) | Industrial product defect detection method based on single sample learning | |
| CN115082650A (en) | Implementation method of automatic pipeline defect labeling tool based on convolutional neural network | |
| CN115564734A (en) | Product detection method, electronic equipment and storage medium | |
| CN114037679A (en) | Product image defect detection method and device based on unsupervised feature combination | |
| CN108763063B (en) | Software defect detection method without defect labeling data | |
| CN114692887A (en) | Semi-supervised learning system and semi-supervised learning method | |
| CN117953206B (en) | Mixed supervision target detection method and device based on point labeling guidance |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20201013 |