CN113344880A - Fast-RCNN-based low-voltage electrical appliance transfer printing pattern defect detection method - Google Patents
Fast-RCNN-based low-voltage electrical appliance transfer printing pattern defect detection method Download PDFInfo
- Publication number
- CN113344880A CN113344880A CN202110643854.1A CN202110643854A CN113344880A CN 113344880 A CN113344880 A CN 113344880A CN 202110643854 A CN202110643854 A CN 202110643854A CN 113344880 A CN113344880 A CN 113344880A
- Authority
- CN
- China
- Prior art keywords
- fast
- rcnn
- layer
- defect detection
- pad printing
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 42
- 230000007547 defect Effects 0.000 title claims abstract description 41
- 238000010023 transfer printing Methods 0.000 title claims description 3
- 238000012549 training Methods 0.000 claims abstract description 28
- 238000007649 pad printing Methods 0.000 claims abstract description 23
- 238000000034 method Methods 0.000 claims abstract description 20
- 238000012545 processing Methods 0.000 claims abstract description 16
- 238000007781 pre-processing Methods 0.000 claims abstract description 11
- 238000011176 pooling Methods 0.000 claims description 24
- 239000013598 vector Substances 0.000 claims description 9
- 230000001629 suppression Effects 0.000 claims description 8
- 238000012360 testing method Methods 0.000 claims description 6
- 238000010586 diagram 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
- 238000004140 cleaning Methods 0.000 claims description 3
- 238000005520 cutting process Methods 0.000 claims description 3
- 238000011478 gradient descent method Methods 0.000 claims description 3
- 238000013507 mapping Methods 0.000 claims description 3
- 238000010845 search algorithm Methods 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 3
- 230000008859 change Effects 0.000 abstract description 5
- 238000013135 deep learning Methods 0.000 abstract description 4
- 230000008569 process Effects 0.000 abstract description 4
- 230000000694 effects Effects 0.000 abstract description 3
- 239000000284 extract Substances 0.000 abstract description 2
- 238000007639 printing Methods 0.000 description 4
- 238000003860 storage Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 230000000630 rising effect Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012546 transfer Methods 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
-
- 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
- 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
- G06T2207/20081—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)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention provides a Fast-RCNN-based low-voltage apparatus pad printing pattern defect detection method, and relates to the technical field of part defect detection and machine vision. The method for detecting the pattern defects of the low-voltage electrical appliance pad printing based on Fast-RCNN comprises the steps of image acquisition, image preprocessing, training set input, image input and result output. The method solves the technical problems that in the prior art, in the process of acquiring the written image by adopting electronic equipment, the vibration of a machine, the change of light and the change of defect positions lead to the fact that a traditional image processing algorithm is difficult to obtain a good detection effect and the phenomena of missing detection and wrong detection often occur. The deep learning algorithm can directly process the original image, extract the characteristics for defect detection, and efficiently apply the network module, so that the detection speed is improved, and the pad printing defect detection performance is improved.
Description
Technical Field
The invention relates to the technical field of part defect detection and machine vision, in particular to a method for detecting a pattern transfer defect of a low-voltage electrical appliance based on Fast-RCNN.
Background
With the continuous progress of science and technology, the production, detection and circulation links of products tend to be automated so as to carry out effective production management and technical improvement. The low-voltage electric appliance may generate various defects in the process of performing written description printing, so that the written description printing of the product is blurred. When a consumer encounters an electric appliance with such defects, product information cannot be clearly and accurately obtained through written printing of the description of the electric appliance. Therefore, before the electric products enter the market, the written printing quality of the low-voltage electric appliance specifications must be detected, and unqualified products must be rejected.
The traditional detection method mainly adopts manual detection, however, in recent years, with the continuous rising of labor cost, the heat tide of robot changing comes along with the rising of the labor cost, and the automatic low-voltage electric appliance pad printing pattern defect detection system has been called.
In the current industrial environment, electronic equipment is adopted in a workshop to acquire written description images, but due to vibration of a machine, change of light and change of defect positions, a traditional image processing algorithm is difficult to obtain a good detection effect, and phenomena such as missing detection, wrong detection and the like often occur.
Disclosure of Invention
The invention aims to provide a Fast-RCNN-based low-voltage electrical appliance pad printing pattern defect detection method, which aims to solve the technical problems that in the prior art, in the process of acquiring an instruction written image by adopting electronic equipment, the traditional image processing algorithm is difficult to obtain a good detection effect and the phenomena of missing detection and wrong detection frequently occur due to vibration of a machine, light change and universal defect positions.
The invention provides a Fast-RCNN-based low-voltage apparatus pad printing pattern defect detection method, which comprises the following steps:
s1, collecting the image
Collecting product information by using electronic equipment, and marking the type of the product information;
s2, image preprocessing
Firstly, cleaning the collected product data information, eliminating misclassified data, and improving the quality of a data set;
then, expanding the size of the data set by adopting a series of image preprocessing methods on all the data to form a training set;
s3, inputting a training set
Sequentially inputting the training set into a multi-task training Fast RCNN network;
s4, inputting image
Inputting images to be recognized of a training set into a network, and obtaining a characteristic diagram through a convolution layer and a pooling layer;
extracting a plurality of candidate frames by adopting a selective search algorithm, finding a feature frame corresponding to each candidate frame in the feature map according to the mapping relation between the candidate frames in the original image and the feature map, and pooling each feature frame to a fixed size in an ROI pooling layer;
the feature frame is processed by a full connection layer to obtain feature vectors with fixed sizes, and the feature vectors are respectively processed by respective average pooling layers to obtain a classification score and a Bbox window regression two output vectors;
s5, outputting the result
And performing non-maximum suppression processing on all data results to generate final target detection and identification results, and finally obtaining a pad printing defect detection result.
Further, in step S1, the electronic device is a CCD camera.
Further, in step S2, the image preprocessing specifically includes:
s201, in the selected image, taking out the target area, and zooming to a fixed size of 224 multiplied by 224; in order to enhance the contrast, the target area is subjected to binarization processing to obtain an original training set, and the training set is processed by the test set in the same way;
s202, carrying out random horizontal turning, random affine transformation, random color dithering and random cutting on the selected image to obtain different sizes and aspect ratios;
the random probability is 0.5-0.8;
and finally, adding the test data into a training set.
Further, in step S3, the structure of the Fast-RCNN network includes: 13 convolutional layers, 4 pooling layers, 1 ROI pooling layer, 2 fully-connected layers and 2 average pooling layers.
Further, in the ROI pooling layer, each feature box is pooled to a fixed size of 5 × 5.
Further, the output of the fully connected layer comprises a cls _ pred layer and a box _ pred layer;
the cls _ pred layer is used for classification;
the box _ pred layer is used to adjust the candidate frame position.
Further, for two branches of the full connection layer, a classification layer and a regression layer of the output layer are trained by using a random gradient descent method until the loss function is converged.
Further, in step S5, the non-maximum suppressing processing of all the results includes: according to the two output branches, non-maximum suppression processing is carried out on each type of defects by using window scores to remove overlapped candidate frames, and finally, a window with the highest classification score of each type is obtained.
Further, in step S5, the defect includes a white block or an edge missing in the pad printing.
The invention provides a Fast-RCNN-based low-voltage apparatus pad printing pattern defect detection method which sequentially comprises the steps of image acquisition, image preprocessing, training set input, image input, result output and the like. The Fast RCNN deep learning method is adopted, multi-task training is achieved, extra feature storage space is not needed, and compared with a traditional template matching method, the Fast RCNN deep learning method improves detection speed and accuracy.
On the basis of a large-scale data set pre-training model, the method extracts the characteristics for detecting the defects through deep learning, has higher learning efficiency, improves the generalization capability of a fault detection model, and reduces the undetected rate and the false detection rate.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
FIG. 1 is a flowchart of a Fast-RCNN-based method for detecting defects in pad-printed patterns of a low-voltage electrical apparatus according to an embodiment of the present invention;
FIG. 2 is a processing flow diagram of a Fast-RCNN-based method for detecting defects in a pad-printed pattern of a low-voltage apparatus according to an embodiment of the present invention.
Detailed Description
The technical solutions of the present invention will be described clearly and completely with reference to the accompanying drawings, and it should be understood that the described embodiments are some, but not all embodiments of the present invention. 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.
In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicate orientations or positional relationships based on the orientations or positional relationships shown in the drawings, and are only for convenience of description and simplicity of description, but do not indicate or imply that the device or element being referred to must have a particular orientation, be constructed and operated in a particular orientation, and thus, should not be construed as limiting the present invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.
In the description of the present invention, it should be noted that, unless otherwise explicitly specified or limited, the terms "mounted," "connected," and "connected" are to be construed broadly, e.g., as meaning either a fixed connection, a removable connection, or an integral connection; can be mechanically or electrically connected; they may be connected directly or indirectly through intervening media, or they may be interconnected between two elements. The specific meanings of the above terms in the present invention can be understood in specific cases to those skilled in the art.
As shown in fig. 1 to 2, in an embodiment of the present invention, a method for detecting a defect of a pad printing pattern of a low voltage electrical apparatus based on Fast-RCNN is provided, which includes the following steps:
s1, collecting the image
Collecting product information by using electronic equipment, and marking the type of the product information;
further, in the above step S1, the electronic device is a CCD camera.
S2, image preprocessing
Firstly, cleaning the collected product data information, eliminating misclassified data, and improving the quality of a data set;
and then, expanding the size of the data set by adopting a series of image preprocessing methods on all the data to form a training set.
Further, in step S2, the image preprocessing specifically includes:
s201, in the selected image, taking out the target area, and zooming to a fixed size of 224 multiplied by 224; in order to enhance the contrast, the target area is subjected to binarization processing to obtain an original training set, and the training set is processed by the test set in the same way;
s202, carrying out random horizontal turning, random affine transformation, random color dithering and random cutting on the selected image to obtain different sizes and aspect ratios;
the random probability is 0.5-0.8;
and finally, adding the test data into a training set.
S3, inputting a training set
And sequentially inputting the training sets into a multi-task training Fast RCNN network.
Further, in the above step S3, the structure of the Fast-RCNN network includes: 13 convolutional layers, 4 pooling layers, 1 ROI pooling layer, 2 fully-connected layers and 2 average pooling layers.
S4, inputting image
Inputting images to be recognized of a training set into a network, and obtaining a characteristic diagram through a convolution layer and a pooling layer;
extracting a plurality of candidate frames by adopting a selective search algorithm, finding a feature frame corresponding to each candidate frame in the feature map according to the mapping relation between the candidate frames in the original image and the feature map, and pooling each feature frame to a fixed size in an ROI pooling layer;
and (4) passing the feature frame through a full connection layer to obtain feature vectors with fixed sizes, and respectively obtaining a classification score and a Bbox window regression two output vectors by the feature vectors passing through respective average pooling layers.
Further, in the above ROI pooling layer, each feature frame is pooled to a fixed size of 5 × 5.
Further, the output of the fully connected layer comprises a cls _ pred layer and a box _ pred layer;
the cls _ pred layer is used for classification;
the box _ pred layer is used to adjust the candidate frame position.
Further, for two branches of the full connection layer, a classification layer and a regression layer of the output layer are trained by using a random gradient descent method until the loss function is converged.
S5, outputting the result
And performing non-maximum suppression processing on all data results to generate final target detection and identification results, and finally obtaining a pad printing defect detection result.
Further, in step S5, the performing of the non-maximum suppression processing on all the results includes: according to the two output branches, non-maximum suppression processing is carried out on each type of defects by using window scores to remove overlapped candidate frames, and finally, a window with the highest classification score of each type is obtained.
Further, in step S5, the defect includes a white block or an edge missing in the pad printing.
It will be understood by those skilled in the art that all or part of the steps of the methods of the above embodiments may be implemented by a program instructing associated hardware, the program may be stored in a computer readable storage medium, and the storage medium includes: ROM, RAM, magnetic disks, optical disks, and the like.
In conclusion, the Fast-RCNN-based method for detecting the pattern defects of the pad printing of the low-voltage electrical appliance has higher learning efficiency and recognition accuracy by directly learning the characteristics which can be used for classification from the collected images, and relieves the detection difficulty caused by factors such as vibration, illumination and the like.
Finally, it should be noted that: the above embodiments are only used to illustrate the technical solution of the present invention, and not to limit the same; while the invention has been described in detail and with reference to the foregoing embodiments, it will be understood by those skilled in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; and the modifications or the substitutions do not make the essence of the corresponding technical solutions depart from the scope of the technical solutions of the embodiments of the present invention.
Claims (9)
1. A method for detecting pattern defects of low-voltage electrical appliance pad printing based on Fast-RCNN is characterized by comprising the following steps:
s1, collecting the image
Collecting product information by using electronic equipment, and marking the type of the product information;
s2, image preprocessing
Firstly, cleaning the collected product data information, eliminating misclassified data, and improving the quality of a data set;
then, expanding the size of the data set by adopting a series of image preprocessing methods on all the data to form a training set;
s3, inputting a training set
Sequentially inputting the training set into a multi-task training Fast RCNN network;
s4, inputting image
Inputting images to be recognized of a training set into a network, and obtaining a characteristic diagram through a convolution layer and a pooling layer;
extracting a plurality of candidate frames by adopting a selective search algorithm, finding a feature frame corresponding to each candidate frame in the feature map according to the mapping relation between the candidate frames in the original image and the feature map, and pooling each feature frame to a fixed size in an ROI pooling layer;
the feature frame is processed by a full connection layer to obtain feature vectors with fixed sizes, and the feature vectors are respectively processed by respective average pooling layers to obtain a classification score and a Bbox window regression two output vectors;
s5, outputting the result
And performing non-maximum suppression processing on all data results to generate final target detection and identification results, and finally obtaining a pad printing defect detection result.
2. The Fast-RCNN-based low voltage electrical pad printing pattern defect detection method of claim 1, wherein in the step S1, the electronic device is a CCD camera.
3. The Fast-RCNN-based low-voltage electrical transfer printing pattern defect detection method of claim 2, wherein in the step S2, the image preprocessing specifically comprises:
s201, in the selected image, taking out the target area, and zooming to a fixed size of 224 multiplied by 224; in order to enhance the contrast, the target area is subjected to binarization processing to obtain an original training set, and the training set is processed by the test set in the same way;
s202, carrying out random horizontal turning, random affine transformation, random color dithering and random cutting on the selected image to obtain different sizes and aspect ratios;
the random probability is 0.5-0.8;
and finally, adding the test data into a training set.
4. The Fast-RCNN-based low-voltage apparatus pad printing pattern defect detection method according to claim 3, wherein in the step S3, the Fast-RCNN network structure comprises: 13 convolutional layers, 4 pooling layers, 1 ROI pooling layer, 2 fully-connected layers and 2 average pooling layers.
5. The Fast-RCNN-based low voltage electrical pad printing pattern defect detection method of claim 4, wherein in the ROI pooling layer each feature box is pooled to a fixed size of 5 x 5.
6. The Fast-RCNN-based low voltage electrical pad printing pattern defect detection method of claim 5, wherein the output of the fully connected layer comprises a cls _ pred layer and a box _ pred layer;
the cls _ pred layer is used for classification;
the box _ pred layer is used to adjust the candidate frame positions.
7. The Fast-RCNN-based low-voltage electrical pad printing pattern defect detection method of claim 6, wherein for both branches of the fully connected layer, the classification layer and the regression layer of the output layer are trained using a stochastic gradient descent method until the loss function converges.
8. The Fast-RCNN-based low-voltage apparatus pad printing pattern defect detection method according to claim 7, wherein in the step S5, the non-maximum suppression processing of all results comprises: according to the two output branches, non-maximum suppression processing is carried out on each type of defects by using window scores to remove overlapped candidate frames, and finally, a window with the highest classification score of each type is obtained.
9. The Fast-RCNN-based low voltage electrical pad printing pattern defect detection method of claim 8, wherein in step S5, the defect comprises a white block or an edge missing in pad printing.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110643854.1A CN113344880A (en) | 2021-06-09 | 2021-06-09 | Fast-RCNN-based low-voltage electrical appliance transfer printing pattern defect detection method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110643854.1A CN113344880A (en) | 2021-06-09 | 2021-06-09 | Fast-RCNN-based low-voltage electrical appliance transfer printing pattern defect detection method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN113344880A true CN113344880A (en) | 2021-09-03 |
Family
ID=77475719
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110643854.1A Pending CN113344880A (en) | 2021-06-09 | 2021-06-09 | Fast-RCNN-based low-voltage electrical appliance transfer printing pattern defect detection method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113344880A (en) |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107301383A (en) * | 2017-06-07 | 2017-10-27 | 华南理工大学 | A kind of pavement marking recognition methods based on Fast R CNN |
CN111126391A (en) * | 2019-12-24 | 2020-05-08 | 熵智科技(深圳)有限公司 | Method for positioning defects of printed characters |
CN111127417A (en) * | 2019-12-20 | 2020-05-08 | 江苏理工学院 | Soft package coil stock printing defect detection method based on SIFT feature matching and improved SSD algorithm |
CN111307835A (en) * | 2020-03-31 | 2020-06-19 | 安徽安视智能科技有限公司 | Printed matter flaw detection method and detection device based on deep learning |
CN211856403U (en) * | 2020-03-31 | 2020-11-03 | 安徽安视智能科技有限公司 | Printed matter flaw detection equipment based on deep learning |
CN112070716A (en) * | 2020-08-03 | 2020-12-11 | 西安理工大学 | Printing defect intelligent identification method based on deep learning |
-
2021
- 2021-06-09 CN CN202110643854.1A patent/CN113344880A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107301383A (en) * | 2017-06-07 | 2017-10-27 | 华南理工大学 | A kind of pavement marking recognition methods based on Fast R CNN |
CN111127417A (en) * | 2019-12-20 | 2020-05-08 | 江苏理工学院 | Soft package coil stock printing defect detection method based on SIFT feature matching and improved SSD algorithm |
CN111126391A (en) * | 2019-12-24 | 2020-05-08 | 熵智科技(深圳)有限公司 | Method for positioning defects of printed characters |
CN111307835A (en) * | 2020-03-31 | 2020-06-19 | 安徽安视智能科技有限公司 | Printed matter flaw detection method and detection device based on deep learning |
CN211856403U (en) * | 2020-03-31 | 2020-11-03 | 安徽安视智能科技有限公司 | Printed matter flaw detection equipment based on deep learning |
CN112070716A (en) * | 2020-08-03 | 2020-12-11 | 西安理工大学 | Printing defect intelligent identification method based on deep learning |
Non-Patent Citations (2)
Title |
---|
AI之路: ""Fast RCNN算法详解"", 《CSDN》, 3 June 2017 (2017-06-03), pages 1 - 14 * |
郑树泉 等, 上海科学技术出版社, pages: 171 - 173 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110070536B (en) | Deep learning-based PCB component detection method | |
CN111784633B (en) | Insulator defect automatic detection algorithm for electric power inspection video | |
CN109615609A (en) | A kind of solder joint flaw detection method based on deep learning | |
CN114240939B (en) | Method, system, equipment and medium for detecting appearance defects of mainboard components | |
CN113205063A (en) | Visual identification and positioning method for defects of power transmission conductor | |
CN110619623B (en) | Automatic identification method for heating of joint of power transformation equipment | |
CN113538375A (en) | PCB defect detection method based on YOLOv5 | |
CN115049908B (en) | Multi-stage intelligent analysis method and system based on embryonic development image | |
CN110008917A (en) | Fault detection method based on the relay protection device constant value forms data that table understands | |
CN114529839A (en) | Unmanned aerial vehicle routing inspection-oriented power transmission line hardware anomaly detection method and system | |
CN111680609B (en) | Foreign matter identification system and method based on image registration and target detection | |
CN112233074A (en) | Power failure detection method based on visible light and infrared fusion image | |
CN115578585A (en) | Industrial image anomaly detection method, system, computer device and storage medium | |
CN113887282A (en) | Detection system and method for any-shape adjacent text in scene image | |
CN113344880A (en) | Fast-RCNN-based low-voltage electrical appliance transfer printing pattern defect detection method | |
CN111091550A (en) | Multi-size self-adaptive PCB solder paste area detection system and detection method | |
CN115937492A (en) | Transformer equipment infrared image identification method based on feature identification | |
CN113033489A (en) | Power transmission line insulator identification and positioning method based on lightweight deep learning algorithm | |
Pérez-Gonzalo et al. | Robust wind turbine blade segmentation from RGB images in the wild | |
CN112287898A (en) | Method and system for evaluating text detection quality of image | |
CN112444522A (en) | Method for detecting defects of insulator string of power system | |
CN111126513A (en) | Universal object real-time learning and recognition system and learning and recognition method thereof | |
CN116485802B (en) | Insulator flashover defect detection method, device, equipment and storage medium | |
Silva et al. | Automatic recognition of electrical grid elements using convolutional neural networks | |
CN117373018A (en) | Method and device for identifying abnormal chip arrangement |
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: 20210903 |
|
RJ01 | Rejection of invention patent application after publication |