CN110826473A - Neural network-based automatic insulator image identification method - Google Patents
Neural network-based automatic insulator image identification method Download PDFInfo
- Publication number
- CN110826473A CN110826473A CN201911060949.XA CN201911060949A CN110826473A CN 110826473 A CN110826473 A CN 110826473A CN 201911060949 A CN201911060949 A CN 201911060949A CN 110826473 A CN110826473 A CN 110826473A
- Authority
- CN
- China
- Prior art keywords
- insulator
- neural network
- image
- steps
- different
- 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
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/20—Scenes; Scene-specific elements in augmented reality scenes
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
- G06V10/462—Salient features, e.g. scale invariant feature transforms [SIFT]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to the technical field of image recognition, in particular to an automatic insulator image recognition method based on a neural network. S1, preprocessing the inspection image data; s2: adding a Dropout feature extraction module; s3: generating a multi-scale feature map by upsampling features of different dimensions; s4: positioning and detecting the multi-scale target; s5: storing the obtained results in different folders in a classified manner; s6: and generating a visual report. The method can automatically process a large amount of shot video data in batches, and timely extract the insulator images by using a neural network method, so that the problems in the videos can be rapidly found in a targeted manner, timely troubleshooting and maintenance are facilitated, the potential safety hazard of power equipment faults is reduced, and the overall safe operation level of a power grid is improved.
Description
Technical Field
The invention relates to the technical field of image recognition, in particular to an automatic insulator image recognition method based on a neural network.
Background
Insulators are important components of a power grid, and in order to ensure safe operation of the power grid, the insulators and other components need to be inspected every year. Unmanned aerial vehicle patrols line is one of the important modes that the power line patrolled and examined. The existing unmanned aerial vehicle line patrol method needs an operator to take a large number of videos by aerial photography, and after the unmanned aerial vehicle is recovered, the massive video data needs to be manually screened, classified and checked. Most recorded content is not valid because of the video specificity, but still requires a little manual effort to see. However, most insulators only appear in a certain frame of image of a video, and more time is spent for workers to find an insulator image in the video content. Therefore, a solution capable of intelligently identifying an insulator image is needed.
Disclosure of Invention
The invention provides an automatic insulator image identification method based on a neural network, aiming at overcoming the problems that more time is spent for finding an insulator image in video content and the efficiency is low in the prior art.
In order to solve the technical problems, the invention adopts the technical scheme that: an insulator image automatic identification method based on a neural network comprises the following steps:
s1, preprocessing the patrol image data;
s2: adding a Dropout feature extraction module;
s3: the preprocessed inspection image data are subjected to up-sampling on features of different dimensions to generate a multi-scale feature map;
s4: carrying out positioning detection on the multi-scale feature map;
s5: storing the obtained results in different folders in a classified manner;
s6: and generating a visual report according to the result of the step S5.
Preferably, the step S1 specifically includes the following steps:
s 11: collecting patrol image data;
s 12: inputting the data of the inspection image into an image identification module, determining whether the code is correct, determining whether the naming format of the inspection image is correct or not, determining the coding format according to the suffix name of the inspection image, and checking whether the inspection image is damaged or not;
s 13: the inspection images are superimposed into a tensor so as to simultaneously predict the improvement efficiency.
Preferably, the step S2 specifically includes the following steps:
s 21: adding Dropout to the convolution network processing step, and adding 5% Dropout to each convolution layer in the feature extraction stage to reduce the detection effect of some objects which are too large in volume and too obvious;
s 22: and residual error unit processing step, namely skipping one layer of linear activation result of a certain layer, and directly adding the linear activation result of the certain layer to the front of the nonlinear activation of the next two layers.
Preferably, the step S3 specifically includes the following steps:
s 31: the characteristics of various dimensions of the insulator are provided at different residual error stages;
s 32: the method comprises the steps of (1) up-sampling the insulator in different scale characteristics;
s 33: and splicing the sampling result and the original features into feature tensors with different scales.
Preferably, the step S4 specifically includes the following steps:
s 41: obtaining prior frames with different sizes for the feature tensor clustering of each scale;
s 42: and directly predicting the relative position of the insulator to obtain the center point of the Bounding Box.
Preferably, the step S5 specifically includes the following steps:
s 51: dividing the data into two types of insulator and insulator-free according to the inspection result of the inspection image data;
s 52: and classifying the images into different categories according to the classification result and storing the images in different folders.
Preferably, the step S6 specifically includes the following steps:
s 71: counting the target object associated with the name of the folder to which the data belongs according to the data;
s 72: and generating a report for subsequent processing by the target species generated by the target statistic result.
Preferably, in the step s11, an unmanned aerial vehicle is used to photograph the insulator image to acquire the inspection image data.
The invention also provides an insulator image automatic identification system based on the neural network, which comprises a memory and a processor, wherein the memory comprises a program of the insulator image automatic identification method based on the neural network, and when the program of the insulator image automatic identification method based on the neural network is executed by the processor, the following steps are realized:
s1, preprocessing the patrol image data;
s2: adding a Dropout feature extraction module;
s3: the preprocessed inspection image data are subjected to up-sampling on features of different dimensions to generate a multi-scale feature map;
s4: carrying out positioning detection on the multi-scale feature map;
s5: storing the obtained results in different folders in a classified manner;
s6: and generating a visual report according to the result of the step S5.
The invention also provides a computer readable storage medium, which contains a program of the method for automatically identifying the insulator image based on the neural network, and when the program of the method for automatically identifying the insulator image based on the neural network is executed by a processor, the steps of the method for automatically identifying the insulator image based on the neural network are realized.
Compared with the prior art, the beneficial effects are:
the method can automatically process a large amount of shot video data in batches, and timely extract the insulator images by using a neural network method, so that the problems in the videos can be rapidly found in a targeted manner, timely troubleshooting and maintenance are facilitated, the potential safety hazard of power equipment faults is reduced, and the overall safe operation level of a power grid is improved. After using patrolling line unmanned aerial vehicle to shoot power equipment, the video of shooing can intelligent processing, carries out work such as insulator extraction and classification for the video. These repeated tasks have conventionally taken a lot of time for power inspectors, and have also caused problems such as false detection and missed detection.
Drawings
Fig. 1 is a flow chart of the method for automatically identifying an insulator image based on a neural network.
Detailed Description
The drawings are for illustrative purposes only and are not to be construed as limiting the patent; for the purpose of better illustrating the embodiments, certain features of the drawings may be omitted, enlarged or reduced, and do not represent the size of an actual product; it will be understood by those skilled in the art that certain well-known structures in the drawings and descriptions thereof may be omitted. The positional relationships depicted in the drawings are for illustrative purposes only and are not to be construed as limiting the present patent.
The same or similar reference numerals in the drawings of the embodiments of the present invention correspond to the same or similar components; in the description of the present invention, it should be understood that if there are terms such as "upper", "lower", "left", "right", "long", "short", etc., indicating orientations or positional relationships based on the orientations or positional relationships shown in the drawings, it is only for convenience of description and simplicity of description, but does not indicate or imply that the device or element referred to must have a specific orientation, be constructed in a specific orientation, and be operated, and therefore, the terms describing the positional relationships in the drawings are only used for illustrative purposes and are not to be construed as limitations of the present patent, and specific meanings of the terms may be understood by those skilled in the art according to specific situations.
The technical scheme of the invention is further described in detail by the following specific embodiments in combination with the attached drawings:
example 1
As shown in fig. 1, an insulator image automatic identification method based on a neural network includes the following steps:
s1, preprocessing the patrol image data;
s2: adding a Dropout feature extraction module;
s3: the preprocessed inspection image data are subjected to up-sampling on features of different dimensions to generate a multi-scale feature map;
s4: carrying out positioning detection on the multi-scale feature map;
s5: storing the obtained results in different folders in a classified manner;
s6: and generating a visual report according to the result of the step S5.
Wherein, step S1 specifically includes the following steps:
s 11: collecting patrol image data;
s 12: inputting the data of the inspection image into an image identification module, determining whether the code is correct, determining whether the naming format of the inspection image is correct or not, determining the coding format according to the suffix name of the inspection image, and checking whether the inspection image is damaged or not;
s 13: the inspection images are superimposed into a tensor so as to simultaneously predict the improvement efficiency.
Further, step S2 specifically includes the following steps:
s 21: adding Dropout to the convolution network processing step, and adding 5% Dropout to each convolution layer in the feature extraction stage to reduce the detection effect of some objects which are too large in volume and too obvious;
s 22: and residual error unit processing step, namely skipping one layer of linear activation result of a certain layer, and directly adding the linear activation result of the certain layer to the front of the nonlinear activation of the next two layers.
Wherein, step S3 specifically includes the following steps:
s 31: the characteristics of various dimensions of the insulator are provided at different residual error stages;
s 32: the method comprises the steps of (1) up-sampling the insulator in different scale characteristics;
s 33: and splicing the sampling result and the original features into feature tensors with different scales.
Further, step S4 specifically includes the following steps:
s 41: obtaining prior frames with different sizes for the feature tensor clustering of each scale;
s 42: and directly predicting the relative position of the insulator to obtain the center point of the Bounding Box.
Wherein, step S5 specifically includes the following steps:
s 51: dividing the data into two types of insulator and insulator-free according to the inspection result of the inspection image data;
s 52: and classifying the images into different categories according to the classification result and storing the images in different folders.
Further, step S6 specifically includes the following steps:
s 71: counting the target object associated with the name of the folder to which the data belongs according to the data;
s 72: and generating a report for subsequent processing by the target species generated by the target statistic result.
Wherein, in step s11, the insulator image is photographed by the unmanned aerial vehicle to collect the inspection image data.
Example 2
An insulator image automatic identification system based on a neural network comprises a memory and a processor, wherein the memory comprises a program of an insulator image automatic identification method based on the neural network, and when the program of the insulator image automatic identification method based on the neural network is executed by the processor, the following steps are realized: s1, preprocessing the patrol image data;
s2: adding a Dropout feature extraction module;
s3: the preprocessed inspection image data are subjected to up-sampling on features of different dimensions to generate a multi-scale feature map;
s4: carrying out positioning detection on the multi-scale feature map;
s5: storing the obtained results in different folders in a classified manner;
s6: and generating a visual report according to the result of the step S5.
Example 3
A computer readable storage medium comprises a program of the neural network-based insulator image automatic identification method, and when the program of the neural network-based insulator image automatic identification method is executed by a processor, the steps of the neural network-based insulator image automatic identification method are realized.
It should be understood that the above-described embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. Other variations and modifications will be apparent to persons skilled in the art in light of the above description. And are neither required nor exhaustive of all embodiments. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present invention should be included in the protection scope of the claims of the present invention.
Claims (10)
1. An insulator image automatic identification method based on a neural network is characterized by comprising the following steps:
s1, preprocessing the patrol image data;
s2: adding a Dropout feature extraction module;
s3: the preprocessed inspection image data are subjected to up-sampling on features of different dimensions to generate a multi-scale feature map;
s4: carrying out positioning detection on the multi-scale feature map;
s5: storing the obtained results in different folders in a classified manner;
s6: and generating a visual report according to the result of the step S5.
2. The method for automatically identifying an insulator image based on a neural network according to claim 1, wherein the step S1 specifically comprises the following steps:
s 11: collecting patrol image data;
s 12: inputting the data of the inspection image into an image identification module, determining whether the code is correct, determining whether the naming format of the inspection image is correct or not, determining the coding format according to the suffix name of the inspection image, and checking whether the inspection image is damaged or not;
s 13: the inspection images are superimposed into a tensor so as to simultaneously predict the improvement efficiency.
3. The method for automatically identifying an insulator image based on a neural network according to claim 1, wherein the step S2 specifically comprises the following steps:
s 21: adding Dropout to the convolution network processing step, and adding 5% Dropout to each convolution layer in the feature extraction stage to reduce the detection effect of some objects which are too large in volume and too obvious;
s 22: and residual error unit processing step, namely skipping one layer of linear activation result of a certain layer, and directly adding the linear activation result of the certain layer to the front of the nonlinear activation of the next two layers.
4. The method for automatically identifying an insulator image based on a neural network according to claim 1, wherein the step S3 specifically comprises the following steps:
s 31: the characteristics of various dimensions of the insulator are provided at different residual error stages;
s 32: the method comprises the steps of (1) up-sampling the insulator in different scale characteristics;
s 33: and splicing the sampling result and the original features into feature tensors with different scales.
5. The method for automatically identifying an insulator image based on a neural network according to claim 1, wherein the step S4 specifically comprises the following steps:
s 41: obtaining prior frames with different sizes for the feature tensor clustering of each scale;
s 42: and directly predicting the relative position of the insulator to obtain the center point of the Bounding Box.
6. The method for automatically identifying an insulator image based on a neural network according to claim 1, wherein the step S5 specifically comprises the following steps:
s 51: dividing the data into two types of insulator and insulator-free according to the inspection result of the inspection image data;
s 52: and classifying the images into different categories according to the classification result and storing the images in different folders.
7. The method for automatically identifying an insulator image based on a neural network according to claim 1, wherein the step S6 specifically comprises the following steps:
s 71: counting the target object associated with the name of the folder to which the data belongs according to the data;
s 72: and generating a report for subsequent processing by the target species generated by the target statistic result.
8. The method for automatically identifying an insulator image based on a neural network according to claim 2, wherein in the step s11, an unmanned aerial vehicle is used for shooting an insulator image to collect patrol inspection image data.
9. An insulator image automatic identification system based on a neural network is characterized by comprising a memory and a processor, wherein the memory comprises a program of an insulator image automatic identification method based on the neural network, and when the program of the insulator image automatic identification method based on the neural network is executed by the processor, the following steps are realized:
s1, preprocessing the patrol image data;
s2: adding a Dropout feature extraction module;
s3: the preprocessed inspection image data are subjected to up-sampling on features of different dimensions to generate a multi-scale feature map;
s4: carrying out positioning detection on the multi-scale feature map;
s5: storing the obtained results in different folders in a classified manner;
s6: and generating a visual report according to the result of the step S5.
10. A computer-readable storage medium, characterized in that the computer-readable storage medium includes a program of a neural network-based insulator image automatic identification method, and when the program of the neural network-based insulator image automatic identification method is executed by a processor, the steps of the neural network-based insulator image automatic identification method according to any one of claims 1 to 7 are implemented.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911060949.XA CN110826473A (en) | 2019-11-01 | 2019-11-01 | Neural network-based automatic insulator image identification method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911060949.XA CN110826473A (en) | 2019-11-01 | 2019-11-01 | Neural network-based automatic insulator image identification method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110826473A true CN110826473A (en) | 2020-02-21 |
Family
ID=69552040
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911060949.XA Pending CN110826473A (en) | 2019-11-01 | 2019-11-01 | Neural network-based automatic insulator image identification method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110826473A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112839213A (en) * | 2021-02-08 | 2021-05-25 | 上海电力大学 | Overhead line insulator fault monitoring system based on 5G communication |
| CN113052104A (en) * | 2021-03-31 | 2021-06-29 | 广东电网有限责任公司 | Insulator positioning and identifying method and system based on image identification |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101271526A (en) * | 2008-04-22 | 2008-09-24 | 深圳先进技术研究院 | Method for object automatic recognition and three-dimensional reconstruction in image processing |
| CN107680090A (en) * | 2017-10-11 | 2018-02-09 | 电子科技大学 | Based on the electric transmission line isolator state identification method for improving full convolutional neural networks |
| CN108509888A (en) * | 2018-03-27 | 2018-09-07 | 百度在线网络技术（北京）有限公司 | Method and apparatus for generating information |
| US20190156202A1 (en) * | 2016-05-02 | 2019-05-23 | Scopito Aps | Model construction in a neural network for object detection |
| CN110033453A (en) * | 2019-04-18 | 2019-07-19 | 国网山西省电力公司电力科学研究院 | Based on the power transmission and transformation line insulator Aerial Images fault detection method for improving YOLOv3 |
-
2019
- 2019-11-01 CN CN201911060949.XA patent/CN110826473A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101271526A (en) * | 2008-04-22 | 2008-09-24 | 深圳先进技术研究院 | Method for object automatic recognition and three-dimensional reconstruction in image processing |
| US20190156202A1 (en) * | 2016-05-02 | 2019-05-23 | Scopito Aps | Model construction in a neural network for object detection |
| CN107680090A (en) * | 2017-10-11 | 2018-02-09 | 电子科技大学 | Based on the electric transmission line isolator state identification method for improving full convolutional neural networks |
| CN108509888A (en) * | 2018-03-27 | 2018-09-07 | 百度在线网络技术（北京）有限公司 | Method and apparatus for generating information |
| CN110033453A (en) * | 2019-04-18 | 2019-07-19 | 国网山西省电力公司电力科学研究院 | Based on the power transmission and transformation line insulator Aerial Images fault detection method for improving YOLOv3 |
Non-Patent Citations (3)
| Title |
|---|
| JOSEPH REDMON ET AL.: "YOLOv3 An Incremental Improvement", 《ARXIV》 * |
| 高强等: "航拍图像中绝缘子目标检测的研究", 《电测与仪表》 * |
| 高志强等: "《深度学习 从入门到实战》", 30 June 2018, 中国铁道出版社 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112839213A (en) * | 2021-02-08 | 2021-05-25 | 上海电力大学 | Overhead line insulator fault monitoring system based on 5G communication |
| CN113052104A (en) * | 2021-03-31 | 2021-06-29 | 广东电网有限责任公司 | Insulator positioning and identifying method and system based on image identification |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108537154B (en) | Power transmission line bird nest identification method based on HOG characteristics and machine learning | |
| CN110705405B (en) | Target labeling method and device | |
| CN111598040B (en) | Construction worker identity recognition and safety helmet wearing detection method and system | |
| CN110084165B (en) | Intelligent identification and early warning method for abnormal events in open scene of power field based on edge calculation | |
| CN111949809B (en) | Intelligent processing method for infrared inspection data of power transmission line | |
| CN111507147A (en) | Intelligent inspection method and device, computer equipment and storage medium | |
| CN109214280B (en) | Shop identification method and device based on street view, electronic equipment and storage medium | |
| CN111709361B (en) | Method for processing inspection data of unmanned aerial vehicle of power transmission line | |
| CN108846331B (en) | Video identification method for judging whether screw fastener of motor train unit chassis falls off or not | |
| US20200074215A1 (en) | Method and system for facilitating detection and identification of vehicle parts | |
| CN112115927A (en) | Intelligent machine room equipment identification method and system based on deep learning | |
| CN110458794B (en) | Quality detection method and device for accessories of rail train | |
| CN112070135A (en) | Power equipment image detection method and device, power equipment and storage medium | |
| CN112287875B (en) | Abnormal license plate recognition method, device, equipment and readable storage medium | |
| WO2019144416A1 (en) | Information processing method and system, cloud processing device and computer program product | |
| CN106371013A (en) | Picture identification-based GIS switch fault automatic identification system | |
| CN110826473A (en) | Neural network-based automatic insulator image identification method | |
| CN113935364A (en) | Unmanned aerial vehicle flame detection method and system based on YOLOv4 network | |
| CN111881984A (en) | Target detection method and device based on deep learning | |
| CN111126112B (en) | Candidate region determination method and device | |
| CN113281780B (en) | Method and device for marking image data and electronic equipment | |
| CN117114420B (en) | Image recognition-based industrial and trade safety accident risk management and control system and method | |
| CN113065462A (en) | Monitoring method, device, equipment and storage medium for power grid overhead line | |
| CN116503807A (en) | Equipment inspection method, device, electronic equipment and computer program product | |
| CN113727022B (en) | Method and device for collecting inspection image, electronic equipment and storage medium |
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: 20200221 |
|
| RJ01 | Rejection of invention patent application after publication |