CN113673333A - Fall detection algorithm in electric power field operation - Google Patents
Fall detection algorithm in electric power field operation Download PDFInfo
- Publication number
- CN113673333A CN113673333A CN202110804299.6A CN202110804299A CN113673333A CN 113673333 A CN113673333 A CN 113673333A CN 202110804299 A CN202110804299 A CN 202110804299A CN 113673333 A CN113673333 A CN 113673333A
- Authority
- CN
- China
- Prior art keywords
- data set
- power field
- training
- total data
- field operation
- 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.)
- Withdrawn
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 26
- 238000012549 training Methods 0.000 claims abstract description 26
- 230000006870 function Effects 0.000 claims abstract description 10
- 238000004519 manufacturing process Methods 0.000 claims abstract description 6
- 238000012544 monitoring process Methods 0.000 claims abstract description 6
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 claims abstract description 5
- 238000013526 transfer learning Methods 0.000 claims abstract description 3
- 238000000034 method Methods 0.000 claims description 6
- 238000010276 construction Methods 0.000 description 3
- 238000013508 migration Methods 0.000 description 2
- 230000005012 migration Effects 0.000 description 2
- 230000006399 behavior Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 230000003111 delayed effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Economics (AREA)
- Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Water Supply & Treatment (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Public Health (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Image Analysis (AREA)
Abstract
A drop detection algorithm in power field operation comprises the following steps: s1: building yolov3 through a python3.6.5+ keras engineering environment, wherein the backbone uses Darknet53, the loss function uses a cross entropy loss function, and the transfer learning is used for loading pre-training parameters; s2: acquiring picture data of a power site to be detected through power site monitoring video or site photographing, marking a safety rope part in the picture data through label, and then manufacturing the picture data into a training set which can be fed into a network; s3: acquiring an open source data set, combining the open source data set and a training set into a total data set, and expanding the total data set in a data enhancement mode; s4: feeding the total data set into a network, then optimizing by using an Adam optimizer, and training by using the total data set to obtain a final model; s5: and detecting the safety rope in the real-time electric power field operation through the model. The drop detection in the power field operation is high in detection speed and accurate in detection result.
Description
Technical Field
The invention relates to the field of electric power field target detection, in particular to a drop detection algorithm in electric power field operation.
Background
With the rapid development of the scale of the power grid, the situations of multiple points, wide range and complex and various operating environments occur in field construction operation. In construction work, unsafe factors of human unsafe behavior, unsafe states of objects and environments always exist objectively, which form risks in production activities and may cause unsafe accidents once out of control.
At present, the occurrence rate of unsafe accidents is reduced by standardization and mechanization of electric power field construction, but the unsafe accidents still occur due to various unsafe factors of electric power field operation, such as falling of objects and falling of people. At the moment, people or objects falling down need to be treated or treated in time, and the damage and the loss are reduced as much as possible. Therefore, the method is vital to quickly and accurately detecting the falling person or object in the power field operation monitoring video, but the existing detection method is low in speed and inaccurate in detection result, and treatment of the falling person or damage of the falling person is delayed.
Therefore, it is necessary to provide a drop detection algorithm in power field operation to overcome the deficiencies of the prior art.
Disclosure of Invention
The invention aims to avoid the defects of the prior art and provides a drop detection algorithm in electric power field operation, which has high detection speed and accurate detection result.
The object of the invention is achieved by the following technical measures.
A drop detection algorithm in power field operation is provided, which comprises the following steps:
s1: yolov3 was constructed through a python3.6.5+ keras engineering environment, where backbone used Darknet53, the loss function used a cross entropy loss function, and the pre-training parameters were loaded using migration learning.
S2: the method comprises the steps of obtaining picture data of a power field to be detected through power field monitoring video or field photographing, marking a safety rope part in the picture data through label limg, and then manufacturing the picture data into a training set which can be fed into a network.
S3: and acquiring an open source data set, combining the open source data set and a training set into a total data set, and expanding the total data set in a data enhancement mode.
S4: and feeding the total data set into a network, then optimizing by using an Adam optimizer, and performing model training by using the total data set to obtain a final model.
S5: and detecting the safety rope in the real-time electric power field operation through the model.
Preferably, the model training in S4 using the total data set is performed by using a decaying learning rate.
Preferably, the model is evaluated by accuracy during the model training process.
The invention relates to a drop detection algorithm in power field operation, which comprises the following steps: s1: building yolov3 through a python3.6.5+ keras engineering environment, wherein the backbone uses Darknet53, the loss function uses a cross entropy loss function, and the transfer learning is used for loading pre-training parameters; s2: acquiring picture data of a power site to be detected through power site monitoring video or site photographing, marking a safety rope part in the picture data through label, and then manufacturing the picture data into a training set which can be fed into a network; s3: acquiring an open source data set, combining the open source data set and a training set into a total data set, and expanding the total data set in a data enhancement mode; s4: feeding the total data set into a network, then optimizing by using an Adam optimizer, and training by using the total data set to obtain a final model; s5: and detecting the safety rope in the real-time electric power field operation through the model. The drop detection in the power field operation is high in detection speed and accurate in detection result, and the safety of the power field operation is improved.
Detailed Description
The invention is further illustrated by the following examples.
Example 1.
A drop detection algorithm in power field operation comprises the following steps: s1: yolov3 was constructed through a python3.6.5+ keras engineering environment, where backbone used Darknet53, the loss function used a cross entropy loss function, and the pre-training parameters were loaded using migration learning. S2: the method comprises the steps of obtaining picture data of a power field to be detected through power field monitoring video or field photographing, marking a safety rope part in the picture data through label limg, and then manufacturing the picture data into a training set which can be fed into a network. S3: and acquiring an open source data set, combining the open source data set and a training set into a total data set, and expanding the total data set in a data enhancement mode. S4: and feeding the total data set into a network, then optimizing by using an Adam optimizer, and performing model training by using the total data set to obtain a final model. S5: and detecting the safety rope in the real-time electric power field operation through the model. In this embodiment, in S4, the model training is performed by using the total data set in a decaying learning rate manner. And evaluating the model through the accuracy in the model training process. The drop detection in the power field operation is high in detection speed and accurate in detection result.
Finally, it should be noted that the above embodiments are only used for illustrating the technical solutions of the present invention and not for limiting the protection scope of the present invention, and although the present invention is described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions can be made on the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.
Claims (3)
1. A drop detection algorithm in power field operation is characterized by comprising the following steps:
s1: building yolov3 through a python3.6.5+ keras engineering environment, wherein the backbone uses Darknet53, the loss function uses a cross entropy loss function, and the transfer learning is used for loading pre-training parameters;
s2: acquiring picture data of a power site to be detected through power site monitoring video or site photographing, marking a falling article part or a falling part of a worker in the picture data through labellimg, and then manufacturing the picture data into a training set which can be fed into a network;
s3: acquiring an open source data set, combining the open source data set and a training set into a total data set, and expanding the total data set in a data enhancement mode;
s4: feeding the total data set into a network, then optimizing by using an Adam optimizer, and performing model training by using the total data set to obtain a final model;
s5: and detecting the safety rope in the real-time electric power field operation through the model.
2. The drop detection algorithm in power field operations according to claim 1, characterized in that: in S4, model training is performed using the total data set in a decaying learning rate manner.
3. The drop detection algorithm in power field operations according to claim 1, characterized in that: and evaluating the model through the accuracy in the model training process.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010796696 | 2020-08-10 | ||
| CN2020107966969 | 2020-08-10 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113673333A true CN113673333A (en) | 2021-11-19 |
Family
ID=78539397
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110804299.6A Withdrawn CN113673333A (en) | 2020-08-10 | 2021-07-16 | Fall detection algorithm in electric power field operation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113673333A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113657165A (en) * | 2020-08-10 | 2021-11-16 | 广东电网有限责任公司 | Dangerous climbing behavior recognition algorithm in electric power field operation |
| CN117824974A (en) * | 2024-03-05 | 2024-04-05 | 深圳市迈腾电子有限公司 | Switch drop test method, device, electronic equipment and computer readable medium |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110910415A (en) * | 2019-11-28 | 2020-03-24 | 重庆中星微人工智能芯片技术有限公司 | Parabolic detection method, device, server and computer readable medium |
| CN111488799A (en) * | 2020-03-13 | 2020-08-04 | 安徽小眯当家信息技术有限公司 | Falling object identification method and system based on image identification |
-
2021
- 2021-07-16 CN CN202110804299.6A patent/CN113673333A/en not_active Withdrawn
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110910415A (en) * | 2019-11-28 | 2020-03-24 | 重庆中星微人工智能芯片技术有限公司 | Parabolic detection method, device, server and computer readable medium |
| CN111488799A (en) * | 2020-03-13 | 2020-08-04 | 安徽小眯当家信息技术有限公司 | Falling object identification method and system based on image identification |
Non-Patent Citations (1)
| Title |
|---|
| 徐桂涛: "煤矿视频监控系统中人员目标跟踪算法的研究与实现", 《中国优秀硕士学位论文全文数据库工程科技Ⅰ辑》, no. 2019, 15 September 2019 (2019-09-15), pages 021 - 360 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113657165A (en) * | 2020-08-10 | 2021-11-16 | 广东电网有限责任公司 | Dangerous climbing behavior recognition algorithm in electric power field operation |
| CN117824974A (en) * | 2024-03-05 | 2024-04-05 | 深圳市迈腾电子有限公司 | Switch drop test method, device, electronic equipment and computer readable medium |
| CN117824974B (en) * | 2024-03-05 | 2024-05-10 | 深圳市迈腾电子有限公司 | Switch drop test method, device, electronic equipment and computer readable medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113673333A (en) | Fall detection algorithm in electric power field operation | |
| CN109726652B (en) | Method for detecting sleeping behavior of person on duty based on convolutional neural network | |
| CN111144263A (en) | Construction worker high-fall accident early warning method and device | |
| CN113673334A (en) | Safety rope detection algorithm in electric power field operation | |
| CN116246407A (en) | Agriculture and forestry area fire early warning supervisory systems based on artificial intelligence | |
| CN113344326A (en) | System and method for identifying dynamic hidden danger and evaluating risk of multi-information fusion special operation site | |
| CN106772656B (en) | A kind of indoor human body detection method based on infrared sensor array | |
| CN112488879A (en) | Intelligent construction site management system based on BIM model | |
| CN113903081A (en) | Visual identification artificial intelligence alarm method and device for images of hydraulic power plant | |
| CN104314599A (en) | Neutral network-based automatic coal drawing system and method for fully mechanized coal mining face | |
| SG10201804179PA (en) | Equipment condition-based corrosion life monitoring system and method | |
| CN103971288B (en) | A kind of Construction of Steel Structure methods of risk assessment based on accident case reasoning | |
| CN110749532A (en) | Hydraulic engineering seepage intelligent monitoring system and method | |
| US11465242B2 (en) | On-line quantitative evaluation method for stability of welding process | |
| CN108205432A (en) | A kind of real-time eliminating method of observation experiment data outliers | |
| CN113657165A (en) | Dangerous climbing behavior recognition algorithm in electric power field operation | |
| CN113673335A (en) | Dangerous boundary-crossing behavior recognition algorithm in electric power field operation | |
| CN106952438A (en) | A kind of fire alarm method based on video image | |
| CN104598970B (en) | Method for detecting work state of climbing frame group | |
| CN112633238A (en) | Electric welding construction detection method based on deep learning image processing | |
| CN115685908A (en) | Electric welding machine welding monitoring method and system based on Internet of things | |
| JP2019200650A5 (en) | ||
| CN115100592A (en) | Method and device for identifying hidden danger of external damage of power transmission channel and storage medium | |
| CN115601010A (en) | Artificial intelligence fire control wisdom information management platform | |
| CN111476922A (en) | Method and device for judging arrival of power transmission line inspection personnel |
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 | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20211119 |