CN110033453A - Based on the power transmission and transformation line insulator Aerial Images fault detection method for improving YOLOv3 - Google Patents
Based on the power transmission and transformation line insulator Aerial Images fault detection method for improving YOLOv3 Download PDFInfo
- Publication number
- CN110033453A CN110033453A CN201910310921.0A CN201910310921A CN110033453A CN 110033453 A CN110033453 A CN 110033453A CN 201910310921 A CN201910310921 A CN 201910310921A CN 110033453 A CN110033453 A CN 110033453A
- Authority
- CN
- China
- Prior art keywords
- insulator
- yolov3
- power transmission
- model
- data set
- 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.)
- Granted
Links
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
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Insulators (AREA)
- Testing Of Short-Circuits, Discontinuities, Leakage, Or Incorrect Line Connections (AREA)
- Electric Cable Installation (AREA)
Abstract
The invention discloses based on the power transmission and transformation line insulator Aerial Images fault detection method for improving YOLOv3, comprising the following steps: S1: establishing insulation Sub Data Set and insulator breakdown data set；S2: using rotation, overturning, the modes such as contrast enhancing carry out data enhancing；S3: using YOLOv3 as basic framework, model modification cut operator is carried out, constructs isolator detecting model；S4: the YOLOv3-tiny network of standard is built；S5: modified YOLOv3 network and YOLOv3-tiny are subjected to model training using insulation Sub Data Set and insulator breakdown data set respectively；S6: two trained models of connection realize the quickly detection end to end to insulator breakdown.Detection method of the invention can fast and accurately recognize the insulator in complex background image and detection and localization is out of order region, raising routing inspection efficiency and the working strength for mitigating patrol officer, the normal operation for electric system escort.
Description
Technical field
The invention belongs to deep learning computer vision fields and electric power defect recognition field, in particular to a kind of to be based on changing
Into the power transmission and transformation line insulator Aerial Images fault detection method of YOLOv3.
Background technique
Insulator is large number of in power transmission and transformation line, plays the important function of electric insulation and mechanical connection.Due to defeated
The physical geographic environment of the various complexity of power transformation line crossing, is chronically exposed under wind and weather solarization, is easy to appear including self-destruction,
The various failures such as flashover.Insulator, which once breaks down, will seriously affect the normal safe transport of electric power, to threaten electricity
The normal operation of Force system, therefore be the important journey of power transmission and transformation line inspection work to the fault detection rapidly and efficiently of insulator
One of sequence.Common routine inspection mode is to shoot inspection image in specific seat in the plane using unmanned plane at this stage, then by ground handling
Personnel carry out fault distinguishing and analysis to inspection image of taking photo by plane by hand, and this mode not only causes the time delay of fault detection,
A large amount of work load is brought for electric inspection process personnel.
For the automation and real-time inspection for realizing power transmission and transformation line, artificial screening can be replaced using image recognition technology
It takes photo by plane inspection picture.But traditional algorithm of target detection is primarily present three point defects, and the feature designed first is low-level feature,
It is insufficient to the expressivity of target；The feature specific aim of its secondary design is single, to power transmission and transformation line target to be detected in complex environment
Under diversity there is no good robustness；Finally for the inspection picture of magnanimity instantly, speed and the equal nothing of detection effect are detected
Method meets practical application in industry demand.
Deep learning is quickly grown since 2012, and the preferable achievement obtained in computer vision field, but at present
The application study that deep learning is used for electric inspection process is less, and is unable to satisfy the real-time and accuracy of power transmission and transformation line inspection
Requirement, therefore, the present invention mainly study by YOLOv3 model be applied to insulator take photo by plane inspection image fault detection with reality
Now efficiently and accurately detects end to end.
Summary of the invention
The purpose of the present invention is to provide based on the power transmission and transformation line insulator Aerial Images fault detection for improving YOLOv3
Method, to solve the problems mentioned in the above background technology.
To achieve the above object, the invention provides the following technical scheme: based on the power transmission and transformation line insulation for improving YOLOv3
Sub- Aerial Images fault detection method, comprising the following steps:
S1: insulation Sub Data Set and insulator breakdown data set are established；
S2: using rotation, overturning, the modes such as contrast enhancing carry out data enhancing；Using LabelImg tool to image
It is marked, obtains corresponding XML file, then by data images and corresponding XML file according to PASCAL VOC data
Collection format carries out tissue storage；
S3: using YOLOv3 as basic framework, progress model modification cut operator, building isolator detecting model, including with
Lower step:
S301: because being fixed seat in the plane shooting inspection picture, insulator can be roughly divided into the larger insulator of close shot shooting
With the relatively pony insulator of vista shot, therefore by YOLOv3 three scales output become two scales output, only retain 13*13 and 26*
26 output scale gives up the output scale of 52*52, and the meter during model inspection is decreased while reducing model size
Calculation amount；
S302: because of a kind of only detection target object of insulator, therefore the overall network of YOLOv3 seems excessively huge, function
Surplus, can carry out model reduction by cut operator, and increase appropriate detects speed；
S303: while carrying out model beta pruning, in order to guarantee not reduce the detection effect of model, by each ruler of script
3 anchor box of degree distribution are changed to 4 anchor box of each scale, and using to hand over and replace standard European than IOU value
The K-means clustering algorithm of distance generates initial 8 anchor box, each scale more anchor box, Ke Yishi
When raising detection effect；
S304:YOLOv3 considers to carry out multi-class division using softmax must satisfy being mutual exclusion between of all categories,
Therefore it abandons softmax function and sigmoid function is used to carry out category division；
S4: the YOLOv3-tiny network of standard is built；
S5: by modified YOLOv3 network and YOLOv3-tiny respectively using insulation Sub Data Set and insulator breakdown
Data set carries out model training；
S6: two trained models of connection realize quickly detecting end to end to insulator breakdown, and step includes:
S601: the various insulation subregions in modified YOLOv3 model inspection image are used；
S602: the insulator area image that will test is sent into tiny-YOLOv3 and carries out fault detection；
S603: failure detection result is reverted in original image, with rectangle frame, fault type and confidence level text mark
Know.
Preferably, insulator data images are to be shot according to inspection specification in specific seat in the plane by unmanned plane in the S1
Collected a plurality of types of insulator Aerial Images of power transmission and transformation line.
It preferably, include revealing in insulator breakdown data set in the S1, the image of a variety of insulator breakdowns such as electric leakage,
Image is the insulator area image containing failure being cropped in initial Aerial Images and keeps original size size.
Preferably, the LabelImg tool in the S2 is a visual image tagged tool, image tagged tool
Not just labelImg, there are also other tools, finally obtained different document form may then converted into model training institute
The data format needed, using needing configuration surroundings python+lxml before the tool, can mark multi-class and can directly generate
The annotation tool of xml document.
Preferably, the priori frame size of 13*13 and 26*26 is to be obtained on coco data set by kmeans algorithm in the S3
The priori frame size arrived.
Preferably, the softmax function in the S304 can be by a K dimensional vector Z containing any real number " compressed " to another
One K is tieed up in real vector σ so that the range of each element is between 0-1, and all elements and be 1.
Preferably, the Sigmoid function in described 304 is often used as the threshold function table of neural network, and variable mappings are arrived
Between 0-1.
Preferably, in the present invention, because isolator detecting model only needs to detect insulator, insulator is not divided specifically
Classification, therefore softmax function is used again.
It compared with prior art, can the beneficial effects of the present invention are: carrying out fault detection using deep neural network model
To realize the real-time inspection of power transmission and transformation line, detection effect is increased substantially with detection speed, improves power transmission and transformation line inspection
Automatization level, reduce the labor intensity of patrol officer, the present invention is targetedly repaired according to specific industrial application scene
Change design neural network structure, so that network model is more targeted, detection effect is preferably ensured, in addition, existing
Part patent of invention only detect to navigate to insulator and be followed by classifier and be directly made whether differentiation for failure, the present invention
It is then directly targeted to trouble location and judges fault type, so that detection is more specialized.
Detailed description of the invention
Fig. 1 is detection method flow chart of the invention；
Fig. 2 is YOLOv3 model structure of the invention；
Fig. 3 is detection process flow chart of the invention.
Specific embodiment
The technical scheme in the embodiments of the invention will be clearly and completely described below, it is clear that described implementation
Example is only a part of the embodiment of the present invention, instead of all the embodiments.Based on the embodiments of the present invention, this field is common
Technical staff's every other embodiment obtained without making creative work belongs to the model that the present invention protects
It encloses.
Embodiment one
Fig. 1-3 is please referred to, the present invention provides a kind of technical solution: based on the power transmission and transformation line insulator boat for improving YOLOv3
Clap image failure detection method, comprising the following steps:
S1: establishing insulation Sub Data Set and insulator breakdown data set, and insulator data images are to be pressed by unmanned plane
According to a plurality of types of insulator Aerial Images of power transmission and transformation line that inspection specification is arrived in specific seat in the plane shooting, collecting, insulator breakdown
Including self-destruction in data set, the image of a variety of insulator breakdowns such as electric leakage, image is to be cropped in initial Aerial Images containing event
The insulator area image of barrier simultaneously keeps original size size；
S2: using rotation, overturning, the modes such as contrast enhancing carry out data enhancing；Using LabelImg tool to image
It is marked, obtains corresponding XML file, then by data images and corresponding XML file according to PASCAL VOC data
Collection format carries out tissue storage, and LabelImg tool is a visual image tagged tool, and image tagged tool is not just
LabelImg, there are also other tools, finally obtained different document form be may then converted into required for model training
Data format, using needing configuration surroundings python+lxml before the tool, can mark multi-class and xml document can be directly generated
Annotation tool；
S3: using YOLOv3 as basic framework, progress model modification cut operator, building isolator detecting model, including with
Lower step:
S301: because being fixed seat in the plane shooting inspection picture, insulator can be roughly divided into the larger insulator of close shot shooting
With the relatively pony insulator of vista shot, therefore by YOLOv3 three scales output become two scales output, only retain 13*13 and 26*
26 output scale gives up the output scale of 52*52, and the meter during model inspection is decreased while reducing model size
The priori frame size of calculation amount, 13*13 and 26*26 are the priori frame size obtained on coco data set by kmeans algorithm；
S302: because of a kind of only detection target object of insulator, therefore the overall network of YOLOv3 seems excessively huge, function
Surplus, can carry out model reduction by cut operator, and increase appropriate detects speed；
S303: while carrying out model beta pruning, in order to guarantee not reduce the detection effect of model, by each ruler of script
3 anchor box of degree distribution are changed to 4 anchor box of each scale, and using to hand over and replace standard European than IOU value
The K-means clustering algorithm of distance generates initial 8 anchor box, each scale more anchor box, Ke Yishi
When raising detection effect；
S304:YOLOv3 considers to carry out multi-class division using softmax must satisfy being mutual exclusion between of all categories,
Therefore abandon softmax function and sigmoid function is used to carry out category division, in the present invention, because of isolator detecting model
It only needs to detect insulator, does not divide insulator classification specifically, therefore use softmax function again, softmax function can incite somebody to action
One K dimensional vector Z containing any real number is tieed up in real vector σ " compressed " to another K, so that the range of each element is in 0-
Between 1, and all elements and be often used as the threshold function table of neural network for 1, Sigmoid function, variable mappings are arrived
Between 0-1；
S4: the YOLOv3-tiny network of standard is built；
S5: by modified YOLOv3 network and YOLOv3-tiny respectively using insulation Sub Data Set and insulator breakdown
Data set carries out model training；
S6: two trained models of connection realize quickly detecting end to end to insulator breakdown, and step includes:
S601: the various insulation subregions in modified YOLOv3 model inspection image are used；
S602: the insulator area image that will test is sent into tiny-YOLOv3 and carries out fault detection；
S603: failure detection result is reverted in original image, with rectangle frame, fault type and confidence level text mark
Know.
Embodiment two
Fig. 1-3 is please referred to, the present invention provides a kind of technical solution: based on the power transmission and transformation line insulator boat for improving YOLOv3
Clap image failure detection method, comprising the following steps:
S1: establishing insulation Sub Data Set and insulator breakdown data set, and insulator data images are to be pressed by unmanned plane
According to a plurality of types of insulator Aerial Images of power transmission and transformation line that inspection specification is arrived in specific seat in the plane shooting, collecting, insulator breakdown
Including self-destruction in data set, the image of a variety of insulator breakdowns such as electric leakage, image is to be cropped in initial Aerial Images containing event
The insulator area image of barrier simultaneously keeps original size size；
S2: using rotation, overturning, the modes such as contrast enhancing carry out data enhancing；Using LabelImg tool to image
It is marked, obtains corresponding XML file, then by data images and corresponding XML file according to PASCAL VOC data
Collection format carries out tissue storage, and LabelImg tool is a visual image tagged tool, and image tagged tool is not just
LabelImg, there are also other tools, finally obtained different document form be may then converted into required for model training
Data format, using needing configuration surroundings python+lxml before the tool, can mark multi-class and xml document can be directly generated
Annotation tool；
S3: using YOLOv3 as basic framework, progress model modification cut operator, building isolator detecting model, including with
Lower step:
S301: because being fixed seat in the plane shooting inspection picture, insulator can be roughly divided into the larger insulator of close shot shooting
With the relatively pony insulator of vista shot, therefore by YOLOv3 three scales output become two scales output, only retain 13*13 and 26*
26 output scale gives up the output scale of 52*52, and the meter during model inspection is decreased while reducing model size
The priori frame size of calculation amount, 13*13 and 26*26 are the priori frame size obtained on coco data set by kmeans algorithm；
S302: because of a kind of only detection target object of insulator, therefore the overall network of YOLOv3 seems excessively huge, function
Surplus, can carry out model reduction by cut operator, and increase appropriate detects speed；
S303: while carrying out model beta pruning, in order to guarantee not reduce the detection effect of model, by each ruler of script
3 anchor box of degree distribution are changed to 6 anchor box of each scale, and using to hand over and replace standard European than IOU value
The K-means clustering algorithm of distance generates initial 12 anchor box, each scale more anchor box, Ke Yishi
When raising detection effect；
S304:YOLOv3 considers to carry out multi-class division using softmax must satisfy being mutual exclusion between of all categories,
Therefore abandon softmax function and sigmoid function is used to carry out category division, in the present invention, because of isolator detecting model
It only needs to detect insulator, does not divide insulator classification specifically, therefore use softmax function again, softmax function can incite somebody to action
One K dimensional vector Z containing any real number is tieed up in real vector σ " compressed " to another K, so that the range of each element is in 0-
Between 1, and all elements and be often used as the threshold function table of neural network for 1, Sigmoid function, variable mappings are arrived
Between 0-1；
S4: the YOLOv3-tiny network of standard is built；
S5: by modified YOLOv3 network and YOLOv3-tiny respectively using insulation Sub Data Set and insulator breakdown
Data set carries out model training；
S6: two trained models of connection realize quickly detecting end to end to insulator breakdown, and step includes:
S601: the various insulation subregions in modified YOLOv3 model inspection image are used；
S602: the insulator area image that will test is sent into tiny-YOLOv3 and carries out fault detection；
S603: failure detection result is reverted in original image, with rectangle frame, fault type and confidence level text mark
Know.
By comparing experiment to above-mentioned two groups of embodiments, embodiment one can detecte with embodiment two to be out of order,
Embodiment one is different from the anchor box that each scale of embodiment two distributes, and the anchor box mistake that embodiment two is distributed
It is more, cause detection to slow, fast without the detection speed of embodiment one, the present invention carries out event using deep neural network model
The real-time inspection of power transmission and transformation line may be implemented in barrier detection, and detection effect is increased substantially with detection speed, improves power transmission and transformation
The automatization level of line data-logging reduces the labor intensity of patrol officer, and the present invention is according to specific industrial application scene, needle
Neural network structure is designed to the modification of property, so that network model is more targeted, detection effect is preferably ensured, this
Outside, existing part patent of invention, which only detects to navigate to insulator and be followed by classifier, is directly made whether sentencing for failure
Not, the present invention is then directly targeted to trouble location and judges fault type, so that detection is more specialized.
It although an embodiment of the present invention has been shown and described, for the ordinary skill in the art, can be with
A variety of variations, modification, replacement can be carried out to these embodiments without departing from the principles and spirit of the present invention by understanding
And modification, the scope of the present invention is defined by the appended.
Claims (8)
1. based on improve YOLOv3 power transmission and transformation line insulator Aerial Images fault detection method, which is characterized in that including with
Lower step:
S1: insulation Sub Data Set and insulator breakdown data set are established；
S2: using rotation, overturning, the modes such as contrast enhancing carry out data enhancing；Image is carried out using LabelImg tool
Label, obtains corresponding XML file, then by data images and corresponding XML file according to PASCAL VOC data set lattice
Formula carries out tissue storage；
S3: using YOLOv3 as basic framework, model modification cut operator is carried out, constructs isolator detecting model, including following step
It is rapid:
S301: because being fixed seat in the plane shooting inspection picture, insulator can be roughly divided into the larger insulator of close shot shooting and remote
The relatively pony insulator of scape shooting, therefore three scales output in YOLOv3 is become into the output of two scales, only retain 13*13's and 26*26
Scale is exported, the output scale of 52*52 is given up, the calculating during model inspection is decreased while reducing model size
Amount；
S302: because of a kind of only detection target object of insulator, therefore the overall network of YOLOv3 seems excessively huge, function mistake
It is surplus, model reduction can be carried out by cut operator, increase appropriate detects speed；
S303: while carrying out model beta pruning, in order to guarantee not reduce the detection effect of model, by each scale of script point
4 anchor box of each scale are changed to 3 anchor box, and using to hand over and replace standard European distance than IOU value
K-means clustering algorithm generate initial 8 anchor box, each more anchor box of scale, can be appropriate
Improve detection effect；
S304:YOLOv3 considers to carry out multi-class division using softmax must satisfy being mutual exclusion between of all categories, therefore puts
It has abandoned softmax function and sigmoid function is used to carry out category division；
S4: the YOLOv3-tiny network of standard is built；
S5: by modified YOLOv3 network and YOLOv3-tiny respectively using insulation Sub Data Set and insulator breakdown data
Collection carries out model training；
S6: two trained models of connection realize quickly detecting end to end to insulator breakdown, and step includes:
S601: the various insulation subregions in modified YOLOv3 model inspection image are used；
S602: the insulator area image that will test is sent into tiny-YOLOv3 and carries out fault detection；
S603: failure detection result is reverted in original image, with rectangle frame, fault type and confidence level words identification.
2. according to as described in claim 1 based on the power transmission and transformation line insulator Aerial Images fault detection side for improving YOLOv3
Method, it is characterised in that: insulator data images are to be shot according to inspection specification in specific seat in the plane by unmanned plane in the S1
Collected a plurality of types of insulator Aerial Images of power transmission and transformation line.
3. according to as described in claim 1 based on the power transmission and transformation line insulator Aerial Images fault detection side for improving YOLOv3
Method, it is characterised in that: it include revealing in insulator breakdown data set in the S1, the image of a variety of insulator breakdowns such as electric leakage,
Image is the insulator area image containing failure being cropped in initial Aerial Images and keeps original size size.
4. according to as described in claim 1 based on the power transmission and transformation line insulator Aerial Images fault detection side for improving YOLOv3
Method, it is characterised in that: the LabelImg tool in the S2 is a visual image tagged tool, and image tagged tool is not
It is only labelImg, there are also other tools, finally obtained different document form be may then converted into needed for model training
The data format wanted, using needing configuration surroundings python+lxml before the tool, can mark multi-class and xml can be directly generated
The annotation tool of file.
5. according to as described in claim 1 based on the power transmission and transformation line insulator Aerial Images fault detection side for improving YOLOv3
Method, it is characterised in that: the priori frame size of 13*13 and 26*26 is to be obtained on coco data set by kmeans algorithm in the S3
The priori frame size arrived.
6. according to as described in claim 1 based on the power transmission and transformation line insulator Aerial Images fault detection side for improving YOLOv3
Method, it is characterised in that: the softmax function in the S304 can be by a K dimensional vector Z containing any real number " compressed " to another
K in a z-axis is tieed up in real vector σ so that the range of each element is between 0-1, and all elements and be 1.
7. according to as described in claim 1 based on the power transmission and transformation line insulator Aerial Images fault detection side for improving YOLOv3
Method, it is characterised in that: the Sigmoid function in described 304 is often used as the threshold function table of neural network, by variable mappings to 0-
Between 1.
8. according to as described in claim 1 based on the power transmission and transformation line insulator Aerial Images fault detection side for improving YOLOv3
Method, it is characterised in that: in the present invention, because isolator detecting model only needs to detect insulator, do not divide insulator specifically
Classification, therefore softmax function is used again.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910310921.0A CN110033453B (en) | 2019-04-18 | 2019-04-18 | Power transmission and transformation line insulator aerial image fault detection method based on improved YOLOv3 |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910310921.0A CN110033453B (en) | 2019-04-18 | 2019-04-18 | Power transmission and transformation line insulator aerial image fault detection method based on improved YOLOv3 |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110033453A true CN110033453A (en) | 2019-07-19 |
| CN110033453B CN110033453B (en) | 2023-02-24 |
Family
ID=67238836
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910310921.0A Active CN110033453B (en) | 2019-04-18 | 2019-04-18 | Power transmission and transformation line insulator aerial image fault detection method based on improved YOLOv3 |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110033453B (en) |
Cited By (27)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110517228A (en) * | 2019-07-30 | 2019-11-29 | 浙江农林大学 | Trunk image rapid detection method based on convolutional neural networks and transfer learning |
| CN110543986A (en) * | 2019-08-27 | 2019-12-06 | 广东电网有限责任公司 | Intelligent monitoring system and monitoring method for external hidden danger of power transmission line |
| CN110705397A (en) * | 2019-09-19 | 2020-01-17 | 东南大学 | YOLOv3 pruning identification insulator defect method suitable for small field sample amount |
| CN110826473A (en) * | 2019-11-01 | 2020-02-21 | 广东电网有限责任公司 | Neural network-based automatic insulator image identification method |
| CN110826520A (en) * | 2019-11-14 | 2020-02-21 | 燕山大学 | Port grab bucket detection method based on improved YOLOv3-tiny algorithm |
| CN110956176A (en) * | 2019-11-14 | 2020-04-03 | 国网山西省电力公司大同供电公司 | Fitting method and detection device for power transmission line between tower foundations |
| CN110992307A (en) * | 2019-11-04 | 2020-04-10 | 华北电力大学（保定） | Insulator positioning and identifying method and device based on YOLO |
| CN111062915A (en) * | 2019-12-03 | 2020-04-24 | 浙江工业大学 | Real-time steel pipe defect detection method based on improved YOLOv3 model |
| CN111160440A (en) * | 2019-12-24 | 2020-05-15 | 广东省智能制造研究所 | Helmet wearing detection method and device based on deep learning |
| CN111239550A (en) * | 2020-02-27 | 2020-06-05 | 东南大学 | Unmanned aerial vehicle full-automatic multi-machine cooperative power transmission line fault identification method for inspection |
| CN111337789A (en) * | 2019-10-23 | 2020-06-26 | 西安科技大学 | Method and system for detecting fault electrical element in high-voltage transmission line |
| CN111582323A (en) * | 2020-04-17 | 2020-08-25 | 山东信通电子股份有限公司 | Power transmission line channel detection method, device and medium |
| CN111753666A (en) * | 2020-05-21 | 2020-10-09 | 西安科技大学 | Method and system for detecting faults of small targets in power transmission line and storage medium |
| CN111784685A (en) * | 2020-07-17 | 2020-10-16 | 国网湖南省电力有限公司 | Power transmission line defect image identification method based on cloud edge cooperative detection |
| CN111784692A (en) * | 2020-08-11 | 2020-10-16 | 国网内蒙古东部电力有限公司 | Method and device for detecting insulator defects in power system and electronic equipment |
| CN112036464A (en) * | 2020-08-26 | 2020-12-04 | 国家电网有限公司 | Insulator infrared image fault detection method based on YOLOv3-tiny algorithm |
| CN112051270A (en) * | 2020-09-10 | 2020-12-08 | 广州云从人工智能技术有限公司 | Power transmission line defect detection method, system, equipment and medium |
| CN112183667A (en) * | 2020-10-31 | 2021-01-05 | 哈尔滨理工大学 | Insulator fault detection method in cooperation with deep learning |
| CN112541389A (en) * | 2020-09-29 | 2021-03-23 | 西安交通大学 | Power transmission line fault detection method based on EfficientDet network |
| CN112598054A (en) * | 2020-12-21 | 2021-04-02 | 福建京力信息科技有限公司 | Power transmission and transformation project quality general-purpose prevention and control detection method based on deep learning |
| CN112614130A (en) * | 2021-01-04 | 2021-04-06 | 东华大学 | Unmanned aerial vehicle power transmission line insulator fault detection method based on 5G transmission and YOLOv3 |
| CN112712516A (en) * | 2021-01-08 | 2021-04-27 | 哈尔滨市科佳通用机电股份有限公司 | High-speed rail bottom rubber strip fault detection method and system based on YOLOv5 |
| CN113112489A (en) * | 2021-04-22 | 2021-07-13 | 池州学院 | Insulator string-dropping fault detection method based on cascade detection model |
| CN113239838A (en) * | 2021-05-24 | 2021-08-10 | 国网湖北省电力有限公司黄石供电公司 | Unmanned aerial vehicle intelligent inspection identification method for power transmission tower |
| CN113297915A (en) * | 2021-04-28 | 2021-08-24 | 江苏师范大学 | Insulator recognition target detection method based on unmanned aerial vehicle inspection |
| CN116721095A (en) * | 2023-08-04 | 2023-09-08 | 杭州瑞琦信息技术有限公司 | Aerial photographing road illumination fault detection method and device |
| US11836968B1 (en) * | 2022-12-08 | 2023-12-05 | Sas Institute, Inc. | Systems and methods for configuring and using a multi-stage object classification and condition pipeline |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| BR102016016197A2 (en) * | 2016-07-12 | 2018-02-06 | Instituto Federal De Educação, Ciência E Tecnologia Do Ceará | INSULATOR FAULT SIGNALING SYSTEM IN HIGH-VOLTAGE STRUCTURES USING ULTRASOUND. |
| CN108527382A (en) * | 2018-04-09 | 2018-09-14 | 上海方立数码科技有限公司 | A kind of crusing robot |
| CN108596886A (en) * | 2018-04-17 | 2018-09-28 | 福州大学 | Aerial Images insulator based on deep learning falls piece fault rapid detecting method |
| CN108629288A (en) * | 2018-04-09 | 2018-10-09 | 华中科技大学 | A kind of gesture identification model training method, gesture identification method and system |
| CN108647655A (en) * | 2018-05-16 | 2018-10-12 | 北京工业大学 | Low latitude aerial images power line foreign matter detecting method based on light-duty convolutional neural networks |
| CN109325454A (en) * | 2018-09-28 | 2019-02-12 | 合肥工业大学 | A kind of static gesture real-time identification method based on YOLOv3 |
-
2019
- 2019-04-18 CN CN201910310921.0A patent/CN110033453B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| BR102016016197A2 (en) * | 2016-07-12 | 2018-02-06 | Instituto Federal De Educação, Ciência E Tecnologia Do Ceará | INSULATOR FAULT SIGNALING SYSTEM IN HIGH-VOLTAGE STRUCTURES USING ULTRASOUND. |
| CN108527382A (en) * | 2018-04-09 | 2018-09-14 | 上海方立数码科技有限公司 | A kind of crusing robot |
| CN108629288A (en) * | 2018-04-09 | 2018-10-09 | 华中科技大学 | A kind of gesture identification model training method, gesture identification method and system |
| CN108596886A (en) * | 2018-04-17 | 2018-09-28 | 福州大学 | Aerial Images insulator based on deep learning falls piece fault rapid detecting method |
| CN108647655A (en) * | 2018-05-16 | 2018-10-12 | 北京工业大学 | Low latitude aerial images power line foreign matter detecting method based on light-duty convolutional neural networks |
| CN109325454A (en) * | 2018-09-28 | 2019-02-12 | 合肥工业大学 | A kind of static gesture real-time identification method based on YOLOv3 |
Non-Patent Citations (7)
| Title |
|---|
| SONG YANAN等: "Rail Surface Defect Detection Method Based on YOLOv3 Deep Learning Networks", 《2018 CHINESE AUTOMATION CONGRESS(CAC)》 * |
| 于科为: "基于卷积神经网络的工件缺陷检测研究", 《信息与电脑(理论版)》 * |
| 吕铄等: "基于改进损失函数的YOLOv3网络", 《计算机系统应用》 * |
| 吴涛等: "轻量级YOLOV3的绝缘子缺陷检测方法", 《计算机工程》 * |
| 戴伟聪等: "遥感图像中飞机的改进YOLOv3实时检测算法", 《光电工程》 * |
| 陈莉君等: "基于深度神经压缩的YOLO加速研究", 《现代计算机》 * |
| 高强等: "航拍图像中绝缘子目标检测的研究", 《电测与仪表》 * |
Cited By (37)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110517228A (en) * | 2019-07-30 | 2019-11-29 | 浙江农林大学 | Trunk image rapid detection method based on convolutional neural networks and transfer learning |
| CN110543986A (en) * | 2019-08-27 | 2019-12-06 | 广东电网有限责任公司 | Intelligent monitoring system and monitoring method for external hidden danger of power transmission line |
| CN110705397A (en) * | 2019-09-19 | 2020-01-17 | 东南大学 | YOLOv3 pruning identification insulator defect method suitable for small field sample amount |
| CN111337789A (en) * | 2019-10-23 | 2020-06-26 | 西安科技大学 | Method and system for detecting fault electrical element in high-voltage transmission line |
| CN110826473A (en) * | 2019-11-01 | 2020-02-21 | 广东电网有限责任公司 | Neural network-based automatic insulator image identification method |
| CN110992307A (en) * | 2019-11-04 | 2020-04-10 | 华北电力大学（保定） | Insulator positioning and identifying method and device based on YOLO |
| CN110956176A (en) * | 2019-11-14 | 2020-04-03 | 国网山西省电力公司大同供电公司 | Fitting method and detection device for power transmission line between tower foundations |
| CN110826520A (en) * | 2019-11-14 | 2020-02-21 | 燕山大学 | Port grab bucket detection method based on improved YOLOv3-tiny algorithm |
| CN110826520B (en) * | 2019-11-14 | 2023-05-26 | 燕山大学 | Port grab bucket detection method based on improved YOLOv3-tiny algorithm |
| CN110956176B (en) * | 2019-11-14 | 2023-06-16 | 国网山西省电力公司大同供电公司 | Method and detection device for fitting transmission line between tower bases |
| CN111062915A (en) * | 2019-12-03 | 2020-04-24 | 浙江工业大学 | Real-time steel pipe defect detection method based on improved YOLOv3 model |
| CN111062915B (en) * | 2019-12-03 | 2023-10-24 | 浙江工业大学 | Real-time steel pipe defect detection method based on improved YOLOv3 model |
| CN111160440A (en) * | 2019-12-24 | 2020-05-15 | 广东省智能制造研究所 | Helmet wearing detection method and device based on deep learning |
| CN111160440B (en) * | 2019-12-24 | 2023-11-21 | 广东省智能制造研究所 | Deep learning-based safety helmet wearing detection method and device |
| CN111239550A (en) * | 2020-02-27 | 2020-06-05 | 东南大学 | Unmanned aerial vehicle full-automatic multi-machine cooperative power transmission line fault identification method for inspection |
| CN111582323A (en) * | 2020-04-17 | 2020-08-25 | 山东信通电子股份有限公司 | Power transmission line channel detection method, device and medium |
| CN111753666A (en) * | 2020-05-21 | 2020-10-09 | 西安科技大学 | Method and system for detecting faults of small targets in power transmission line and storage medium |
| CN111753666B (en) * | 2020-05-21 | 2024-01-23 | 西安科技大学 | Small target fault detection method, detection system and storage medium for power transmission line |
| CN111784685B (en) * | 2020-07-17 | 2023-08-18 | 国网湖南省电力有限公司 | Power transmission line defect image identification method based on cloud edge cooperative detection |
| CN111784685A (en) * | 2020-07-17 | 2020-10-16 | 国网湖南省电力有限公司 | Power transmission line defect image identification method based on cloud edge cooperative detection |
| CN111784692A (en) * | 2020-08-11 | 2020-10-16 | 国网内蒙古东部电力有限公司 | Method and device for detecting insulator defects in power system and electronic equipment |
| CN112036464A (en) * | 2020-08-26 | 2020-12-04 | 国家电网有限公司 | Insulator infrared image fault detection method based on YOLOv3-tiny algorithm |
| CN112051270A (en) * | 2020-09-10 | 2020-12-08 | 广州云从人工智能技术有限公司 | Power transmission line defect detection method, system, equipment and medium |
| CN112541389A (en) * | 2020-09-29 | 2021-03-23 | 西安交通大学 | Power transmission line fault detection method based on EfficientDet network |
| CN112541389B (en) * | 2020-09-29 | 2023-08-22 | 西安交通大学 | Transmission line fault detection method based on EfficientDet network |
| CN112183667A (en) * | 2020-10-31 | 2021-01-05 | 哈尔滨理工大学 | Insulator fault detection method in cooperation with deep learning |
| CN112598054B (en) * | 2020-12-21 | 2023-09-22 | 福建京力信息科技有限公司 | Power transmission and transformation project quality common disease prevention and detection method based on deep learning |
| CN112598054A (en) * | 2020-12-21 | 2021-04-02 | 福建京力信息科技有限公司 | Power transmission and transformation project quality general-purpose prevention and control detection method based on deep learning |
| CN112614130A (en) * | 2021-01-04 | 2021-04-06 | 东华大学 | Unmanned aerial vehicle power transmission line insulator fault detection method based on 5G transmission and YOLOv3 |
| CN112712516B (en) * | 2021-01-08 | 2021-10-15 | 哈尔滨市科佳通用机电股份有限公司 | High-speed rail bottom rubber strip fault detection method and system based on YOLOv5 |
| CN112712516A (en) * | 2021-01-08 | 2021-04-27 | 哈尔滨市科佳通用机电股份有限公司 | High-speed rail bottom rubber strip fault detection method and system based on YOLOv5 |
| CN113112489A (en) * | 2021-04-22 | 2021-07-13 | 池州学院 | Insulator string-dropping fault detection method based on cascade detection model |
| CN113297915A (en) * | 2021-04-28 | 2021-08-24 | 江苏师范大学 | Insulator recognition target detection method based on unmanned aerial vehicle inspection |
| CN113239838A (en) * | 2021-05-24 | 2021-08-10 | 国网湖北省电力有限公司黄石供电公司 | Unmanned aerial vehicle intelligent inspection identification method for power transmission tower |
| US11836968B1 (en) * | 2022-12-08 | 2023-12-05 | Sas Institute, Inc. | Systems and methods for configuring and using a multi-stage object classification and condition pipeline |
| CN116721095A (en) * | 2023-08-04 | 2023-09-08 | 杭州瑞琦信息技术有限公司 | Aerial photographing road illumination fault detection method and device |
| CN116721095B (en) * | 2023-08-04 | 2023-11-03 | 杭州瑞琦信息技术有限公司 | Aerial photographing road illumination fault detection method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110033453B (en) | 2023-02-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110033453A (en) | Based on the power transmission and transformation line insulator Aerial Images fault detection method for improving YOLOv3 | |
| CN112884931B (en) | Unmanned aerial vehicle inspection method and system for transformer substation | |
| CN108537154B (en) | Power transmission line bird nest identification method based on HOG characteristics and machine learning | |
| CN110703800A (en) | Unmanned aerial vehicle-based intelligent identification method and system for electric power facilities | |
| CN106356757A (en) | Method for inspecting electric power lines by aid of unmanned aerial vehicle on basis of human vision characteristics | |
| CN108648169A (en) | The method and device of high voltage power transmission tower defects of insulator automatic identification | |
| Wang et al. | Research on image recognition of insulators based on YOLO algorithm | |
| CN112101088A (en) | Automatic unmanned aerial vehicle power inspection method, device and system | |
| CN111244822B (en) | Fixed-wing unmanned aerial vehicle line patrol method, system and device in complex geographic environment | |
| CN109389322A (en) | The disconnected broken lot recognition methods of grounded-line based on target detection and long memory models in short-term | |
| CN115240093B (en) | Automatic power transmission channel inspection method based on visible light and laser radar point cloud fusion | |
| CN112367400A (en) | Intelligent inspection method and system for power internet of things with edge cloud coordination | |
| CN112455676A (en) | Intelligent monitoring and analyzing system and method for health state of photovoltaic panel | |
| CN105303162A (en) | Target proposed algorithm-based insulator recognition algorithm for aerial images | |
| CN116258980A (en) | Unmanned aerial vehicle distributed photovoltaic power station inspection method based on vision | |
| CN110490261B (en) | Positioning method for power transmission line inspection image insulator | |
| CN113252701A (en) | Cloud edge cooperation-based power transmission line insulator self-explosion defect detection system and method | |
| CN113534188B (en) | Tower deformation defect detection method based on unmanned aerial vehicle laser point cloud modeling | |
| CN108470141B (en) | Statistical feature and machine learning-based insulator identification method in distribution line | |
| CN114167245B (en) | Intelligent detection method for partial discharge on surface of power transmission and transformation equipment and unmanned aerial vehicle fusion ultraviolet system | |
| CN113361520B (en) | Transmission line equipment defect detection method based on sample offset network | |
| Yao et al. | Research on technology of autonomous inspection system for UAV based on improved YOLOv4 | |
| CN112508905A (en) | Hardware rust image detection method and computer readable storage medium | |
| Zan et al. | Defect Identification of Power Line Insulator Based on an Improved yolov4-tiny Algorithm | |
| CN112132826A (en) | Pole tower accessory defect inspection image troubleshooting method and system based on artificial intelligence |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |