CN110059076A - A kind of Mishap Database semi-automation method for building up of power transmission and transformation line equipment - Google Patents
A kind of Mishap Database semi-automation method for building up of power transmission and transformation line equipment Download PDFInfo
- Publication number
- CN110059076A CN110059076A CN201910315522.3A CN201910315522A CN110059076A CN 110059076 A CN110059076 A CN 110059076A CN 201910315522 A CN201910315522 A CN 201910315522A CN 110059076 A CN110059076 A CN 110059076A
- Authority
- CN
- China
- Prior art keywords
- data
- power transmission
- transformation line
- line equipment
- model
- 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
- 230000005540 biological transmission Effects 0.000 title claims abstract description 55
- 230000009466 transformation Effects 0.000 title claims abstract description 55
- 238000000034 method Methods 0.000 title claims abstract description 31
- 238000001514 detection method Methods 0.000 claims abstract description 33
- 238000013135 deep learning Methods 0.000 claims abstract description 22
- 238000003860 storage Methods 0.000 claims abstract description 20
- 238000013480 data collection Methods 0.000 claims abstract description 11
- 238000012360 testing method Methods 0.000 claims description 29
- 238000004458 analytical method Methods 0.000 claims description 19
- 238000012549 training Methods 0.000 claims description 18
- 230000011514 reflex Effects 0.000 claims description 4
- 238000012216 screening Methods 0.000 claims description 4
- 230000008859 change Effects 0.000 claims description 3
- 238000005457 optimization Methods 0.000 abstract description 12
- 230000010354 integration Effects 0.000 abstract description 3
- 238000010276 construction Methods 0.000 abstract 1
- 238000007689 inspection Methods 0.000 description 7
- 230000008569 process Effects 0.000 description 5
- 230000000694 effects Effects 0.000 description 4
- 230000006872 improvement Effects 0.000 description 4
- 239000000463 material Substances 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 238000013024 troubleshooting Methods 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 230000005611 electricity Effects 0.000 description 2
- 230000004888 barrier function Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 230000001934 delay Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000012212 insulator Substances 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000000116 mitigating effect Effects 0.000 description 1
- 238000004064 recycling Methods 0.000 description 1
- 238000009666 routine test Methods 0.000 description 1
- 230000006641 stabilisation Effects 0.000 description 1
- 238000011105 stabilization Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/21—Design, administration or maintenance of databases
- G06F16/211—Schema design and management
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/21—Design, administration or maintenance of databases
- G06F16/215—Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors
-
- 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
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/20—Administration of product repair or maintenance
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/07—Target detection
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- General Health & Medical Sciences (AREA)
- General Business, Economics & Management (AREA)
- Molecular Biology (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- Software Systems (AREA)
- Marketing (AREA)
- Mathematical Physics (AREA)
- Computing Systems (AREA)
- Quality & Reliability (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Primary Health Care (AREA)
- Entrepreneurship & Innovation (AREA)
- Water Supply & Treatment (AREA)
- Public Health (AREA)
- Operations Research (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of Mishap Database semi-automation method for building up of power transmission and transformation line equipment, the following steps are included: acquisition data: first with the related data of data acquisition device acquisition power transmission and transformation line equipment, and data are screened, non-classified database is established.A kind of Mishap Database semi-automation method for building up of power transmission and transformation line equipment of the present invention, solves the automatic classification of power transmission and transformation line equipment and Mishap Database first, detection and storage problem, it completes power transmission and transformation line equipment and work is established in Mishap Database semi-automation, secondly by the work such as model training-intelligent classification-intelligence test-model optimization integration, form complete closed loop, construct the Database Systems that can independently optimize, it finally can be quick, it is efficient to establish the power transmission and transformation line key equipment and fault data collection for being suitable for deep learning, solves the problems, such as dataset construction hardly possible, bring better prospect of the application.
Description
Technical field
The present invention relates to electric system and computer vision field, in particular to a kind of number of faults of power transmission and transformation line equipment
According to library semi-automation method for building up.
Background technique
Important infrastructure of the power transmission and transformation line as power industry, is the important component of power grid, and transmission line of electricity closes
Key member mainly includes insulator, fitting and shaft tower etc., if component goes wrong, can jeopardize the stabilization of entire operation of power networks
Property, deep learning since two thousand six, can be described as having led the scientific and technological revolution under a big data era to a certain extent,
2012, even more achieve extraordinary effect in image recognition and target detection, currently, deep learning recognition of face,
In particular kind of target detection, it is significantly better than conventional method, recognition effect is very good, but in concrete engineering application: for example defeated
On power transformation line facility and fault detection, recognition effect just differs greatly.
Its basic reason is, lacks power transmission and transformation line key equipment and fault data collection, currently, International Publication
In the data sets such as PASCAL VOC, ImageNet, MS-COCO, Open Images Dataset, Sun, without disclosed defeated change
Electric line key equipment and fault data collection, this problem have caused great difficulties to the research in the field, and in production number
During collection, there is also following two points problems:
1. manual sort's data, take time and effort, after data acquisition, data format, generally photo or video, and collection process
In not can guarantee the high quality of content of shooting, need manually to carry out preliminary quality screening, secondly, data acquisition content is multiple portions
Part, and manually when to single goal component trouble-shooting, it needs first to carry out manual sort to all data, then be directed to specific objective
Component carries out trouble shoot work, and whole process consumes a large amount of human and material resources and financial resources, seriously delays job schedule；
2. markers work amount is big, for single component or fault data collection, none complete data analysing method is led
It causes to exist in data set and largely influences small data on model, so that the picture number for needing to mark is big, increase label
The workload of work extends data set time, wastes a large amount of manpowers；
3. working for model optimization, there is also model test results using difficult problem, and model optimization work, is depth
Learning objective detects basic, starts with from model test results, analyzes undetected image, with model and Data set reconstruction two
A direction is pushed further into, and could complete model optimization work, and during Data set reconstruction, it needs to analyze and not detected
The characteristics of image arrived finds data again from all data sets, after pretreatment, reconstructed data set in proportion, to optimize mould
Type, this process need to analyse in depth characteristics of image, and repeated data is found and markers work, works and increases to model optimization
Difficulty.
Therefore, a kind of power transmission and transformation line equipment based on deep learning and Mishap Database semi-automation foundation side are invented
Method has very important practical value, significant to the intelligentized promotion of power transmission and transformation line failure inspection work.
Summary of the invention
The main purpose of the present invention is to provide a kind of Mishap Database semi-automation foundation sides of power transmission and transformation line equipment
Method can effectively solve the problems in background technique.
To achieve the above object, the technical scheme adopted by the invention is as follows:
A kind of Mishap Database semi-automation method for building up of power transmission and transformation line equipment, comprising the following steps:
(1), data are acquired: acquiring the related data of power transmission and transformation line equipment, and logarithm first with data acquisition device
According to being screened, non-classified database is established；
(2), data are analyzed: analysis obtains the characteristic information of target component, picks out characteristic information obviously and picture quality
High partial data carries out target component and marks work, and mark file and picture name correspond；
(3), data classification: being based on deep learning target detection principle, the data set that will have been marked, according to the ratio of 8:2,
It is randomly divided into training set and test set, carries out model training work；
(4), optimal models are found: model training being carried out to two group models, a kind of model, which uses, combines region
The R-CNN list of target detection model of proposal and CNN network, mainly include R-CNN, SPP-Net, Fast R-CNN,
Faster R-CNN and R-FCN, another model use the model that target detection is converted to regression problem, mainly include YOLO
Series and SSD；
(5), analysis model test result: choosing whole result optimal models, to the unused data of residue and Xin Cai
The data of collection carry out intelligent classification storage and work with fault detection, and result is entered into taxonomy database system；
(6), it establishes data set: utilizing the classification results of step (5), repeat step (2), establish and be suitable for deep learning
Power transmission and transformation line key equipment and fault data collection；
(7), result detects: inspecting periodically classification storage as a result, the data for not meeting model to result analyse in depth its figure
It as feature, and finds data similar with its feature and carries out Data set reconstruction, adjust model parameter, and then Optimized model.
Preferably, in the step (1), data acquisition device is using unmanned plane load high-definition camera or telephoto lens list
Instead, with the mode of data flow when data acquire, batch obtains power transmission and transformation line equipment and equipment fault picture and video.
Preferably, in the step (2), the high standard of picture quality is that picture pixels are greater than 6,000,000, picture without ghost image and
Have no occluder.
Preferably, in the step (3), data set according to the ratio of 7~9:1~3 can also classify.
Preferably, in the step (5), accuracy rate, the recall rate, AP of test result are analyzed when analysis model test result
Value, mAP value and test result picture.
Preferably, in the step (7), characteristics of image includes but is not limited to the color characteristic, textural characteristics, shape of image
Shape feature and spatial relation characteristics.
Preferably, in the step (7), the time for inspecting periodically classification storage result is 3-15 days.
Compared with prior art, the invention has the following beneficial effects:
1, the workload for mitigating power transmission and transformation line inspection staff significantly, using small part data, training pattern, to
Screening, the classification work of the following most of data of detection, save the consumption of a large amount of human and material resources, save the cost；
2, using deep learning target detection principle, keep electric inspection process work intelligent, pass through trained power transmission and transformation line
Road key equipment fault model, automatically to total data carry out fault identification and storage work, one it is found that, overhaul immediately, solve
Automatic classification, detection and the storage problem of power transmission and transformation line equipment of having determined and Mishap Database, complete power transmission and transformation line equipment and
Work is established in Mishap Database semi-automation, compared to the method for traditional artificial searching, reduces the time of trouble-shooting, allows electricity
Power inspection work is efficient, while substantially reducing the security risk of power grid；
3, by the work such as model training-intelligent classification-intelligence test-model optimization integration, complete closed loop is formed, is constructed
Intelligent classification practical, optimization property is strong, a fault-detection data library system out；
4, it is the electric system power transmission and transformation line research work based on deep learning, data support is provided, notebook data is utilized
The data classified in library can fast and efficiently establish the power transmission and transformation line key equipment suitable for deep learning and event
Hinder data set, is the optimization and improvement of successive depths learning model algorithm, data set is provided.
Detailed description of the invention
Fig. 1 is a kind of holistic approach stream of the Mishap Database semi-automation method for building up of power transmission and transformation line equipment of the present invention
Cheng Tu；
Fig. 2 is a kind of overall flow of the Mishap Database semi-automation method for building up of power transmission and transformation line equipment of the present invention
Figure.
Specific embodiment
To be easy to understand the technical means, the creative features, the aims and the efficiencies achieved by the present invention, below with reference to
Specific embodiment, the present invention is further explained.
Embodiment 1
(1), data are acquired: acquiring the related data of power transmission and transformation line equipment, and logarithm first with data acquisition device
According to being screened, non-classified database is established, data acquisition device is using unmanned plane load high-definition camera or telephoto lens
Single-lens reflex camera, with the mode of data flow when data acquire, batch obtains power transmission and transformation line equipment and equipment fault picture and video；
(2), data are analyzed: analysis obtains the characteristic information of target component, picks out characteristic information obviously and picture quality
High partial data carries out target component and marks work, and mark file and picture name correspond, and the high standard of picture quality is
Picture pixels are greater than 6,000,000, and picture is without ghost image and have no occluder；
(3), data classification: being based on deep learning target detection principle, the data set that will have been marked, according to the ratio of 7:3,
It is randomly divided into training set and test set, carries out model training work；
(4), optimal models are found: model training being carried out to two group models, a kind of model, which uses, combines region
The R-CNN list of target detection model of proposal and CNN network, mainly include R-CNN, SPP-Net, Fast R-CNN,
Faster R-CNN and R-FCN, another model use the model that target detection is converted to regression problem, mainly include YOLO
Series and SSD；
(5), analysis model test result: choosing whole result optimal models, to the unused data of residue and Xin Cai
The data of collection carry out intelligent classification storage and work with fault detection, and result is entered into taxonomy database system, analyze mould
Accuracy rate, recall rate, AP value, mAP value and the test result picture of test result are analyzed when type test result；
(6), it establishes data set: utilizing the classification results of step (5), repeat step (2), establish and be suitable for deep learning
Power transmission and transformation line key equipment and fault data collection；
(7), result detects: inspecting periodically classification storage as a result, the data for not meeting model to result analyse in depth its figure
It as feature, and finds data similar with its feature and carries out Data set reconstruction, adjust model parameter, and then Optimized model, periodically
The time for checking classification storage result is 5 days, and characteristics of image includes but is not limited to the color characteristic, textural characteristics, shape of image
Shape feature and spatial relation characteristics.
Embodiment 2
(1), data are acquired: acquiring the related data of power transmission and transformation line equipment, and logarithm first with data acquisition device
According to being screened, non-classified database is established, data acquisition device is using unmanned plane load high-definition camera or telephoto lens
Single-lens reflex camera, with the mode of data flow when data acquire, batch obtains power transmission and transformation line equipment and equipment fault picture and video；
(2), data are analyzed: analysis obtains the characteristic information of target component, picks out characteristic information obviously and picture quality
High partial data carries out target component and marks work, and mark file and picture name correspond, and the high standard of picture quality is
Picture pixels are greater than 6,000,000, and picture is without ghost image and have no occluder；
(3), data classification: being based on deep learning target detection principle, the data set that will have been marked, according to the ratio of 8:2,
It is randomly divided into training set and test set, carries out model training work；
(4), optimal models are found: model training being carried out to two group models, a kind of model, which uses, combines region
The R-CNN list of target detection model of proposal and CNN network, mainly include R-CNN, SPP-Net, Fast R-CNN,
Faster R-CNN and R-FCN, another model use the model that target detection is converted to regression problem, mainly include YOLO
Series and SSD；
(5), analysis model test result: choosing whole result optimal models, to the unused data of residue and Xin Cai
The data of collection carry out intelligent classification storage and work with fault detection, and result is entered into taxonomy database system, analyze mould
Accuracy rate, recall rate, AP value, mAP value and the test result picture of test result are analyzed when type test result；
(6), it establishes data set: utilizing the classification results of step (5), repeat step (2), establish and be suitable for deep learning
Power transmission and transformation line key equipment and fault data collection；
(7), result detects: inspecting periodically classification storage as a result, the data for not meeting model to result analyse in depth its figure
It as feature, and finds data similar with its feature and carries out Data set reconstruction, adjust model parameter, and then Optimized model, periodically
The time for checking classification storage result is 8 days, and characteristics of image includes but is not limited to the color characteristic, textural characteristics, shape of image
Shape feature and spatial relation characteristics.
Embodiment 3
(1), data are acquired: acquiring the related data of power transmission and transformation line equipment, and logarithm first with data acquisition device
According to being screened, non-classified database is established, data acquisition device is using unmanned plane load high-definition camera or telephoto lens
Single-lens reflex camera, with the mode of data flow when data acquire, batch obtains power transmission and transformation line equipment and equipment fault picture and video；
(2), data are analyzed: analysis obtains the characteristic information of target component, picks out characteristic information obviously and picture quality
High partial data carries out target component and marks work, and mark file and picture name correspond, and the high standard of picture quality is
Picture pixels are greater than 6,000,000, and picture is without ghost image and have no occluder；
(3), data classification: being based on deep learning target detection principle, the data set that will have been marked, according to the ratio of 9:1,
It is randomly divided into training set and test set, carries out model training work；
(4), optimal models are found: model training being carried out to two group models, a kind of model, which uses, combines region
The R-CNN list of target detection model of proposal and CNN network, mainly include R-CNN, SPP-Net, Fast R-CNN,
Faster R-CNN and R-FCN, another model use the model that target detection is converted to regression problem, mainly include YOLO
Series and SSD；
(5), analysis model test result: choosing whole result optimal models, to the unused data of residue and Xin Cai
The data of collection carry out intelligent classification storage and work with fault detection, and result is entered into taxonomy database system, analyze mould
Accuracy rate, recall rate, AP value, mAP value and the test result picture of test result are analyzed when type test result；
(6), it establishes data set: utilizing the classification results of step (5), repeat step (2), establish and be suitable for deep learning
Power transmission and transformation line key equipment and fault data collection；
(7), result detects: inspecting periodically classification storage as a result, the data for not meeting model to result analyse in depth its figure
It as feature, and finds data similar with its feature and carries out Data set reconstruction, adjust model parameter, and then Optimized model, periodically
The time for checking classification storage result is 12 days, and characteristics of image includes but is not limited to the color characteristic, textural characteristics, shape of image
Shape feature and spatial relation characteristics.
It should be noted that a kind of skill of the Mishap Database semi-automation method for building up of power transmission and transformation line equipment of the present invention
Art scheme is to utilize the part power transmission and transformation line equipment and fault data collection marked based on deep learning target detection principle
Training objective detection model chooses test result optimal models, act on be not used and freshly harvested power transmission and transformation line equipment and
Failure picture, to be classified, be detected and be stored work, routine test model analyses in depth test result, and recycling has number
According to reconstructed data set, the Optimization Work of propulsion model, so that database semi-automation method for building up is completed, meanwhile, utilize the party
Method, repeat mark work, can fast and efficiently establish the power transmission and transformation line key equipment and fault data suitable for deep learning
Collection；
Using this method, the workload of power transmission and transformation line inspection staff can be mitigated significantly, using small part data,
Training pattern saves the consumption of a large amount of human and material resources, saves to detect screening, the classification work of following most of data
Cost；
Using deep learning target detection principle, keeps electric inspection process work intelligent, pass through trained power transmission and transformation line
Key equipment fault model, automatically to total data carry out fault identification and storage work, one it is found that, overhaul immediately, solve
Automatic classification, detection and the storage problem of power transmission and transformation line equipment and Mishap Database, complete power transmission and transformation line equipment and therefore
Work is established in barrier database semi-automation, compared to the method for traditional artificial searching, is reduced the time of trouble-shooting, is allowed electric power
Inspection work is efficient, while substantially reducing the security risk of power grid；
The work such as model training-intelligent classification-intelligence test-model optimization integration can also be formed complete closed loop, structure
Produce intelligent classification practical, optimization property is strong, a fault-detection data library system；
It finally can also be the electric system power transmission and transformation line research work based on deep learning, data support, benefit are provided
With the data classified in database, can fast and efficiently establish crucial suitable for the power transmission and transformation line of deep learning
Equipment and fault data collection are the optimization and improvement of successive depths learning model algorithm, provide data set, using effect phase
It is more preferable for traditional approach, meet the use demand of people.
The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention.The technology of the industry
Personnel are it should be appreciated that the present invention is not limited to the above embodiments, and the above embodiments and description only describe this
The principle of invention, without departing from the spirit and scope of the present invention, various changes and improvements may be made to the invention, these changes
Change and improvement all fall within the protetion scope of the claimed invention.The claimed scope of the invention by appended claims and its
Equivalent thereof.
Claims (7)
1. a kind of Mishap Database semi-automation method for building up of power transmission and transformation line equipment, comprising the following steps:
(1), acquire data: first with data acquisition device acquisition power transmission and transformation line equipment related data, and to data into
Row screening, establishes non-classified database；
(2), data are analyzed: analysis obtains the characteristic information of target component, picks out that characteristic information is obvious and picture quality is high
Partial data carries out target component and marks work, and mark file and picture name correspond；
(3), data classification: based on deep learning target detection principle, the data set that will have been marked, according to the ratio of 8:2, at random
It is divided into training set and test set, carries out model training work；
(4), optimal models are found: model training carried out to two group models, a kind of model use combine region proposal and
The R-CNN list of target detection model of CNN network mainly includes R-CNN, SPP-Net, Fast R-CNN, Faster R-CNN
And R-FCN, another model use the model that target detection is converted to regression problem, mainly include YOLO series and SSD；
(5), analysis model test result: choosing whole result optimal models, to unused data of residue and freshly harvested
Data carry out intelligent classification storage and work with fault detection, and result is entered into taxonomy database system；
(6), it establishes data set: utilizing the classification results of step (5), repeat step (2), establish the defeated change for being suitable for deep learning
Electric line key equipment and fault data collection；
(7), result detects: inspecting periodically classification storage as a result, the data for not meeting model to result analyse in depth its image spy
Sign, and find data similar with its feature and carry out Data set reconstruction, adjust model parameter, and then Optimized model.
2. a kind of Mishap Database semi-automation method for building up of power transmission and transformation line equipment according to claim 1, special
Sign is: in the step (1), data acquisition device is using unmanned plane load high-definition camera or telephoto lens single-lens reflex camera, data
With the mode of data flow when acquisition, batch obtains power transmission and transformation line equipment and equipment fault picture and video.
3. a kind of Mishap Database semi-automation method for building up of power transmission and transformation line equipment according to claim 1, special
Sign is: in the step (2), the high standard of picture quality is that picture pixels are greater than 6,000,000, and picture is without ghost image and unobstructed
Object.
4. a kind of Mishap Database semi-automation method for building up of power transmission and transformation line equipment according to claim 1, special
Sign is: in the step (3), data set can also classify according to the ratio of 7~9:1~3.
5. a kind of Mishap Database semi-automation method for building up of power transmission and transformation line equipment according to claim 1, special
Sign is: in the step (5), accuracy rate, recall rate, AP value, the mAP value of test result are analyzed when analysis model test result
With test result picture.
6. a kind of Mishap Database semi-automation method for building up of power transmission and transformation line equipment according to claim 1, special
Sign is: in the step (7), characteristics of image include but is not limited to the color characteristic of image, textural characteristics, shape feature and
Spatial relation characteristics.
7. a kind of Mishap Database semi-automation method for building up of power transmission and transformation line equipment according to claim 1, special
Sign is: in the step (7), the time for inspecting periodically classification storage result is 3-15 days.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910315522.3A CN110059076A (en) | 2019-04-19 | 2019-04-19 | A kind of Mishap Database semi-automation method for building up of power transmission and transformation line equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910315522.3A CN110059076A (en) | 2019-04-19 | 2019-04-19 | A kind of Mishap Database semi-automation method for building up of power transmission and transformation line equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110059076A true CN110059076A (en) | 2019-07-26 |
Family
ID=67319638
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910315522.3A Pending CN110059076A (en) | 2019-04-19 | 2019-04-19 | A kind of Mishap Database semi-automation method for building up of power transmission and transformation line equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110059076A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110597874A (en) * | 2019-08-26 | 2019-12-20 | 数字广东网络建设有限公司 | Data analysis model creation method and device, computer equipment and storage medium |
| CN111815581A (en) * | 2020-06-27 | 2020-10-23 | 国网上海市电力公司 | Power isolating switch and automatic part identification method and device thereof |
| CN112102443A (en) * | 2020-09-15 | 2020-12-18 | 国网电力科学研究院武汉南瑞有限责任公司 | Marking system and marking method suitable for substation equipment inspection image |
| CN113984018A (en) * | 2021-10-27 | 2022-01-28 | 南方电网数字电网研究院有限公司 | Site digital standard investigation method based on mobile phone positioning |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107704878A (en) * | 2017-10-09 | 2018-02-16 | 南京大学 | A kind of high-spectral data storehouse semi-automation method for building up based on deep learning |
| CN108257114A (en) * | 2017-12-29 | 2018-07-06 | 天津市万贸科技有限公司 | A kind of transmission facility defect inspection method based on deep learning |
| US20180260687A1 (en) * | 2016-04-26 | 2018-09-13 | Hitachi, Ltd. | Information Processing System and Method for Operating Same |
| CN108761237A (en) * | 2018-05-29 | 2018-11-06 | 福州大学 | Unmanned plane electric inspection process image vital electrical component diagnoses automatically and labeling system |
| CN109389322A (en) * | 2018-10-30 | 2019-02-26 | 福州大学 | The disconnected broken lot recognition methods of grounded-line based on target detection and long memory models in short-term |
-
2019
- 2019-04-19 CN CN201910315522.3A patent/CN110059076A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180260687A1 (en) * | 2016-04-26 | 2018-09-13 | Hitachi, Ltd. | Information Processing System and Method for Operating Same |
| CN107704878A (en) * | 2017-10-09 | 2018-02-16 | 南京大学 | A kind of high-spectral data storehouse semi-automation method for building up based on deep learning |
| CN108257114A (en) * | 2017-12-29 | 2018-07-06 | 天津市万贸科技有限公司 | A kind of transmission facility defect inspection method based on deep learning |
| CN108761237A (en) * | 2018-05-29 | 2018-11-06 | 福州大学 | Unmanned plane electric inspection process image vital electrical component diagnoses automatically and labeling system |
| CN109389322A (en) * | 2018-10-30 | 2019-02-26 | 福州大学 | The disconnected broken lot recognition methods of grounded-line based on target detection and long memory models in short-term |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110597874A (en) * | 2019-08-26 | 2019-12-20 | 数字广东网络建设有限公司 | Data analysis model creation method and device, computer equipment and storage medium |
| CN110597874B (en) * | 2019-08-26 | 2022-05-27 | 数字广东网络建设有限公司 | Data analysis model creation method and device, computer equipment and storage medium |
| CN111815581A (en) * | 2020-06-27 | 2020-10-23 | 国网上海市电力公司 | Power isolating switch and automatic part identification method and device thereof |
| CN112102443A (en) * | 2020-09-15 | 2020-12-18 | 国网电力科学研究院武汉南瑞有限责任公司 | Marking system and marking method suitable for substation equipment inspection image |
| CN113984018A (en) * | 2021-10-27 | 2022-01-28 | 南方电网数字电网研究院有限公司 | Site digital standard investigation method based on mobile phone positioning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110059076A (en) | A kind of Mishap Database semi-automation method for building up of power transmission and transformation line equipment | |
| Li et al. | Sewer pipe defect detection via deep learning with local and global feature fusion | |
| Wang et al. | Recognition and drop-off detection of insulator based on aerial image | |
| CN111754498A (en) | Conveyor belt carrier roller detection method based on YOLOv3 | |
| CN107194396A (en) | Method for early warning is recognized based on the specific architecture against regulations in land resources video monitoring system | |
| CN111814678A (en) | Video monitoring-based method and system for identifying coal flow in conveyor belt | |
| CN112270681B (en) | Method and system for detecting and counting yellow plate pests deeply | |
| CN110245592A (en) | A method of for promoting pedestrian's weight discrimination of monitoring scene | |
| CN113205039A (en) | Power equipment fault image identification and disaster investigation system and method based on multiple DCNNs | |
| CN114359167B (en) | Insulator defect detection method based on lightweight YOLOv < 4 > under complex scene | |
| Chen et al. | An improved faster R-CNN transmission line bolt defect detection method | |
| CN114359578A (en) | Application method and system of pest and disease damage identification intelligent terminal | |
| CN113361520B (en) | Transmission line equipment defect detection method based on sample offset network | |
| Terzoglou et al. | Employing deep learning framework for improving solar panel defects using drone imagery | |
| Sharma et al. | Object detection in power line infrastructure: A review of the challenges and solutions | |
| CN113139476A (en) | Data center-oriented human behavior attribute real-time detection method and system | |
| CN113469938A (en) | Pipe gallery video analysis method and system based on embedded front-end processing server | |
| CN117456384A (en) | High-precision unmanned aerial vehicle electric power inspection system | |
| Zhou | A lightweight improvement of YOLOv5 for insulator fault detection | |
| CN101876993A (en) | Method for extracting and retrieving textural features from ground digital nephograms | |
| Zhang et al. | Unmanned Aerial Vehicle (UAV) vision-based detection of power line poles by CPU-based deep learning method | |
| CN108470141B (en) | Statistical feature and machine learning-based insulator identification method in distribution line | |
| Liu et al. | Abnormal behavior analysis strategy of bus drivers based on deep learning | |
| CN111783891B (en) | Customized object detection method | |
| Luo et al. | Hardhat-wearing detection with cloud-edge collaboration in Power Internet-of-Things |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190726 |