CN113762339A - Safety helmet detection algorithm in electric power field operation - Google Patents
Safety helmet detection algorithm in electric power field operation Download PDFInfo
- Publication number
- CN113762339A CN113762339A CN202110883056.6A CN202110883056A CN113762339A CN 113762339 A CN113762339 A CN 113762339A CN 202110883056 A CN202110883056 A CN 202110883056A CN 113762339 A CN113762339 A CN 113762339A
- Authority
- CN
- China
- Prior art keywords
- safety helmet
- training
- data set
- model
- helmet detection
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 51
- 238000012549 training Methods 0.000 claims abstract description 64
- 238000000034 method Methods 0.000 claims abstract description 12
- 238000012360 testing method Methods 0.000 claims abstract description 10
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 claims abstract description 5
- 238000002372 labelling Methods 0.000 claims description 6
- 230000006399 behavior Effects 0.000 description 4
- 238000010276 construction Methods 0.000 description 4
- 238000012544 monitoring process Methods 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 230000005684 electric field Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000012423 maintenance Methods 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
- 230000009466 transformation Effects 0.000 description 1
Images
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
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Helmets And Other Head Coverings (AREA)
Abstract
A safety helmet detection algorithm in power field operation comprises the following steps: s1: constructing a training data set; s2: deleting the samples with missing values in the training data set by a dropna method; s3: expanding the training data set in a data enhancement mode; s4: encoding the extended training data set; s5: dividing samples in the coded training data set into a training set and a test set; s6: constructing a safety helmet detection model, comprising: s61: configuring a caffe environment, and installing a faster rcnn model; s62: leading the samples in the training set into a fast rcnn model for training to obtain a safety helmet detection model; s7: introducing the sample concentrated in the test into a safety helmet detection model, collecting the recognition accuracy of the safety helmet detection model, and optimizing the safety helmet detection model by using an Adam optimizer; s8: and detecting real-time picture data in the power field operation through the optimized safety helmet detection model. The safety helmet in the electric power field operation has high and accurate detection and identification rate.
Description
Technical Field
The invention relates to the technical field of safety identification, in particular to a safety helmet detection algorithm in electric power field operation.
Background
The whole of the substation and the transmission and distribution line of various voltages in the power system is called a power grid. The system comprises three units of power transformation, power transmission and power distribution. The task of the power grid is to deliver and distribute electrical energy, changing the voltage.
The installation and maintenance of the power system are generally referred to as a power construction site. The electric power construction site has complex environment and multiple points, and needs to pay attention to unsafe behaviors of workers on the construction site all the time. At present, unsafe behaviors of people in an electric power construction site are mainly detected by using a camera for real-time monitoring, monitoring workers mainly detect safety helmets worn by the workers, and the positions of the workers are detected in real time by detecting the safety helmets to judge safety conditions of the workers. However, the existing helmet detection method through the camera only adopts artificial detection through monitoring, so that the detection accuracy is low, unsafe behaviors of workers cannot be completely detected, and some unsafe accidents can still occur.
Therefore, it is necessary to provide a safety helmet 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 safety helmet detection algorithm in electric field operation, which has high recognition rate and accurate recognition.
The object of the invention is achieved by the following technical measures.
The provided safety helmet detection algorithm in power field operation comprises the following steps:
s1: constructing a training data set;
specifically, acquiring picture data of a power site, labeling a safety cap part of each picture in the picture data, and preparing the picture data into a training set which can be fed into a network;
acquiring an open source data set, and combining the open source data set with a data set which can be fed into a network to construct a training data set;
s2: deleting the samples with missing values in the training data set by a dropna method;
s3: expanding the training data set in a data enhancement mode;
s4: encoding the extended training data set;
s5: dividing samples in the coded training data set into a training set and a test set;
s6: the method for constructing the safety helmet detection model specifically comprises the following steps:
s61: configuring a caffe environment, and installing a faster rcnn model;
s62: leading the samples in the training set into a faster rcnn model for training, wherein the obtained model is a safety helmet detection model;
s7: introducing the sample concentrated in the test into a safety helmet detection model, collecting the recognition accuracy of the safety helmet detection model, and optimizing the safety helmet detection model by using an Adam optimizer;
s8: and detecting real-time picture data in the power field operation through the optimized safety helmet detection model.
Preferably, the model training in step S62 is performed using a training set in a learning rate reduction manner.
Preferably, the model is evaluated by accuracy during the model training process.
Preferably, labeler performs labeler labeling in step S1 using labellimg.
The invention relates to a safety helmet detection algorithm in power field operation, which comprises the following steps: s1: constructing a training data set; specifically, picture data of a power site are obtained, a safety cap part of each picture in the picture data is labeled, and then the picture data are prepared into a training set which can be fed into a network. Acquiring an open source data set, and combining the open source data set with a data set which can be fed into a network to construct a training data set; s2: deleting the samples with missing values in the training data set by a dropna method; s3: expanding the training data set in a data enhancement mode; s4: encoding the extended training data set; s5: dividing samples in the coded training data set into a training set and a test set; s6: the method for constructing the safety helmet detection model specifically comprises the following steps: s61: configuring a caffe environment, and installing a faster rcnn model; s62: leading the samples in the training set into a faster rcnn model for training, wherein the obtained model is a safety helmet detection model; s7: introducing the sample concentrated in the test into a safety helmet detection model, collecting the recognition accuracy of the safety helmet detection model, and optimizing the safety helmet detection model by using an Adam optimizer; s8: and detecting real-time picture data in the power field operation through the optimized safety helmet detection model. The safety helmet in the electric power field operation has high and accurate detection and identification rate, and is favorable for improving the safety of the electric power field operation.
Drawings
Fig. 1 is one of the picture samples of the training data set in example 1.
Fig. 2 is one of the picture samples of the training data set in example 1.
Fig. 3 is one of the picture samples of the training data set in example 1.
Detailed Description
The invention is further illustrated by the following examples.
Example 1.
A safety helmet detection algorithm in power field operation comprises the following steps:
s1: constructing a training data set;
specifically, acquiring picture data of a power site, labeling a safety cap part of each picture in the picture data, and preparing the picture data into a training set which can be fed into a network; and acquiring an open source data set, and combining the open source data set with a data set which can be fed into a network to construct a training data set. And one part of the data in the training data set is from the Internet, the other part of the data is shot by the user, and the picture data in the training data set are picture samples which are labeled on the safety helmet. The picture samples in this example are shown in figures 1, 2 and 3.
S2: deleting the samples with missing values in the training data set by a dropna method;
s3: expanding the training data set in a data enhancement mode;
s4: encoding the extended training data set;
s5: dividing samples in the coded training data set into a training set and a test set;
s6: the method for constructing the safety helmet detection model specifically comprises the following steps:
s61: configuring a caffe environment, and installing a faster rcnn model;
s62: leading the samples in the training set into a faster rcnn model for training, wherein the obtained model is a safety helmet detection model;
s7: introducing the sample concentrated in the test into a safety helmet detection model, collecting the recognition accuracy of the safety helmet detection model, and optimizing the safety helmet detection model by using an Adam optimizer;
s8: and detecting real-time picture data in the power field operation through the optimized safety helmet detection model.
The safety helmet in the electric power field operation has high detection and calculation recognition rate and high accuracy, and can effectively avoid unsafe behaviors in the electric power field operation.
Example 2.
The other characteristics of the safety helmet detection algorithm in the electric field operation are the same as those of the embodiment 1, and the difference is that: in step S62, model training is performed using the training set in a learning rate reduction manner. And evaluating the model through accuracy in the model training process. In step S1, labeler is used to perform labeling.
The safety helmet in the electric power field operation has high detection and calculation recognition rate and high accuracy, and model training is performed in a mode of attenuating the learning rate, so that the model training can be accelerated, and the model is close to the optimal model.
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 (4)
1. A safety helmet detection algorithm in power field operation is characterized by comprising the following steps:
s1: constructing a training data set;
specifically, acquiring picture data of a power site, labeling a safety cap part of each picture in the picture data, and preparing the picture data into a training set which can be fed into a network;
acquiring an open source data set, and combining the open source data set with a data set which can be fed into a network to construct a training data set;
s2: deleting the samples with missing values in the training data set by a dropna method;
s3: expanding the training data set in a data enhancement mode;
s4: encoding the extended training data set;
s5: dividing samples in the coded training data set into a training set and a test set;
s6: the method for constructing the safety helmet detection model specifically comprises the following steps:
s61: configuring a caffe environment, and installing a faster rcnn model;
s62: leading the samples in the training set into a faster rcnn model for training, wherein the obtained model is a safety helmet detection model;
s7: introducing the sample concentrated in the test into a safety helmet detection model, collecting the recognition accuracy of the safety helmet detection model, and optimizing the safety helmet detection model by using an Adam optimizer;
s8: and detecting real-time picture data in the power field operation through the optimized safety helmet detection model.
2. The safety helmet detection algorithm in power field operations of claim 1, wherein: in step S62, model training is performed using the training set in a learning rate reduction manner.
3. The safety helmet detection algorithm in power field operations of claim 1, wherein: and evaluating the model through accuracy in the model training process.
4. The safety helmet detection algorithm in power field operations of claim 1, wherein: in step S1, labeler is used to perform labeling.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110883056.6A CN113762339A (en) | 2021-08-02 | 2021-08-02 | Safety helmet detection algorithm in electric power field operation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110883056.6A CN113762339A (en) | 2021-08-02 | 2021-08-02 | Safety helmet detection algorithm in electric power field operation |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113762339A true CN113762339A (en) | 2021-12-07 |
Family
ID=78788361
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110883056.6A Pending CN113762339A (en) | 2021-08-02 | 2021-08-02 | Safety helmet detection algorithm in electric power field operation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113762339A (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110738127A (en) * | 2019-09-19 | 2020-01-31 | 福建师范大学福清分校 | Helmet identification method based on unsupervised deep learning neural network algorithm |
| CN111914743A (en) * | 2020-07-31 | 2020-11-10 | 广东电网有限责任公司清远供电局 | Method and device for detecting safety helmet of transformer substation worker |
-
2021
- 2021-08-02 CN CN202110883056.6A patent/CN113762339A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110738127A (en) * | 2019-09-19 | 2020-01-31 | 福建师范大学福清分校 | Helmet identification method based on unsupervised deep learning neural network algorithm |
| CN111914743A (en) * | 2020-07-31 | 2020-11-10 | 广东电网有限责任公司清远供电局 | Method and device for detecting safety helmet of transformer substation worker |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110018389B (en) | Online fault monitoring method and system for power transmission line | |
| CN108537394B (en) | Real-time safety early warning method and device for smart power grid | |
| CN108268706B (en) | Closed loop test system and method for intelligent alarm application of integrated monitoring system | |
| CN104410163B (en) | A kind of safety in production based on electric energy management system and power-economizing method | |
| CN110969194B (en) | Cable early fault positioning method based on improved convolutional neural network | |
| CN107730117B (en) | Cable maintenance early warning method and system based on heterogeneous data comprehensive analysis | |
| CN104504525A (en) | Method for realizing power-grid equipment failure prewarning through big data mining technology | |
| CN109655712A (en) | A kind of distribution network line fault analysis of causes method and system | |
| CN106779095B (en) | Intelligent substation equipment soft pressing plate checking method based on KMP algorithm | |
| CN104700226A (en) | Equipment and device for monitoring environmental protection measures in transmission and transformation project construction process | |
| CN110197475A (en) | Insulator automatic recognition system, method and application in a kind of transmission line of electricity | |
| CN110165674B (en) | Active filter safety management system | |
| CN112085233A (en) | Power digital information model based on station domain BIM data fusion multi-source information | |
| CN104156888A (en) | Power system operation risk monitoring method based on comprehensive risk evaluation model | |
| CN113673334A (en) | Safety rope detection algorithm in electric power field operation | |
| CN113762339A (en) | Safety helmet detection algorithm in electric power field operation | |
| CN116545111A (en) | Information interaction system based on internal and external networks of transformer substation | |
| CN110310048B (en) | Distribution network planning overall process evaluation method and device | |
| CN104573366A (en) | Method for verifying electric quantity data based on calculation of electric power load integrals | |
| CN117092953A (en) | Production data acquisition management and control system based on industrial Internet of things | |
| CN116566839A (en) | Communication resource quality evaluation system for power enterprises | |
| CN108573233B (en) | Power grid ceramic insulator identification method based on image processing technology | |
| CN110955808A (en) | Method for matching and associating four remote monitoring information of main plant station | |
| CN103390035A (en) | Intelligent warning signal type matching method based on regular expressions | |
| CN116170283A (en) | Processing method based on network communication fault system |
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 |