CN112232133A - Power transmission line image identification method and device based on deep convolutional neural network - Google Patents
Power transmission line image identification method and device based on deep convolutional neural network Download PDFInfo
- Publication number
- CN112232133A CN112232133A CN202010989285.1A CN202010989285A CN112232133A CN 112232133 A CN112232133 A CN 112232133A CN 202010989285 A CN202010989285 A CN 202010989285A CN 112232133 A CN112232133 A CN 112232133A
- Authority
- CN
- China
- Prior art keywords
- transmission line
- power transmission
- image
- neural network
- convolutional neural
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/182—Network patterns, e.g. roads or rivers
-
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
- G06F18/24133—Distances to prototypes
- G06F18/24137—Distances to cluster centroïds
- G06F18/2414—Smoothing the distance, e.g. radial basis function networks [RBFN]
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computational Linguistics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Evolutionary Biology (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to a method and a device for identifying an image of a power transmission line based on a deep convolutional neural network, which comprises the steps of preprocessing the image of the power transmission line as a sample set of model training, uniformly processing the sizes of pictures of fault sample pictures such as outburst fault pictures, forest fire fault pictures and the like, marking the types of the faults and generating a marked file set; secondly, training a sample picture set by adopting a deep convolution neural network method based on feature map extraction to generate a model file; and finally, deploying an image recognition device based on a deep convolutional neural network model at the front end of the camera to realize the rapid and accurate recognition of the transmission line fault. The power transmission line image identification method and device based on the feature map deep convolutional neural network can improve timeliness of image identification of multiple fault types and reduce consumption rate of network resources and server memory resources for image identification of images transmitted back to a background server.
Description
Technical Field
The invention relates to the technical field of image processing, in particular to a power transmission line image identification method and device based on a deep convolutional neural network.
Background
The transmission line is a main network of the power system, has large scale, long distance and wide distribution region, is generally distributed in remote zones, and has severe operation environment. Natural disasters such as mountain fires, landslides and the like damage the normal operation of the overhead transmission line, and meanwhile, the danger of external force damage caused by construction machinery, ultrahigh vehicles and ships exists. The traditional mode of adopting artifical line inspection has the problem that intensity of labour is big, work efficiency is low, and it is poor to the sudden natural disasters and the outer broken incident of coping with the timeliness. At present, the safe operation condition of the power transmission line is sensed in real time by adopting a video and image remote monitoring method, and the intelligent monitoring system has the characteristics of high intelligent level, strong timeliness and the like, and is an important component for constructing an intelligent power grid.
However, the intelligent operation and maintenance monitoring platform of the power transmission line generates a large amount of video and image information every day, and the monitoring platform is only relied on to identify faults, so that the operating pressure of the server is high, and the capacity expansion cost of the server is increased. In addition, the network resource occupation rate is large and the communication cost is high due to the fact that all image information is transmitted to the server side.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides the power transmission line image identification method and the power transmission line image identification device based on the deep convolutional neural network, which can improve the timeliness of image identification of multiple fault types and reduce the consumption rate of network resources and server memory resources for image identification when the images are transmitted back to a background server.
In order to achieve the purpose, the invention provides a power transmission line image identification method based on a deep convolutional neural network, which comprises the following steps:
(1) preprocessing the image of the power transmission line, marking the fault type and the coordinate position to form a sample set of model training;
(2) establishing a deep convolution neural network image recognition model based on feature map extraction, and training and packaging by adopting the samples in the sample set;
(3) and collecting the image of the power transmission line, identifying by adopting the packaged deep convolutional neural network image identification model, and judging whether a fault occurs and the fault position.
Further, the preprocessing the image of the power transmission line and the marking of the fault type and the coordinate position in the step (1) comprise:
screening the power transmission line images generated by the power transmission line monitoring system by adopting a clustering analysis method, and selecting various fault type image samples;
carrying out size processing on all the electric transmission line pictures by utilizing a Python programming language to enable the electric transmission line pictures to become standard sizes input by model training, naming and marking fault types by adopting a uniform picture naming rule;
and marking the fault positions of various fault types in the power transmission line image by using an image marking software tool to generate a sample marking file.
Further, the fault types comprise power transmission line construction machinery, a tower crane, a crane and a mountain fire.
Further, in the step (2), a deep convolutional neural network image recognition model based on feature map extraction is established, and the following steps are performed by adopting the samples in the sample set for training and encapsulating:
constructing a deep convolutional neural network algorithm model based on feature map extraction by adopting a deep learning method of single-step classification identification;
adopting the samples in the sample set as the input of a deep convolutional neural network model, setting the learning rate, the variation factor and the model training times in a network model configuration file, and carrying out deep convolutional neural network model training until the precision meets the requirement or the training times are finished;
selecting a test sample from the samples in the sample set, and identifying accuracy precision of a deep convolutional neural network model of the test sample;
adjusting the learning rate, the variation factor and the model training times, retraining the deep convolutional neural network model, and testing the recognition accuracy of the deep convolutional neural network model;
comparing the accuracy rates of the two deep convolutional neural network models, selecting the model file of the deep convolutional neural network model with higher identification accuracy rate, and realizing model optimization.
Another aspect of the present invention provides a power transmission line image recognition apparatus based on a deep convolutional neural network, including: the system comprises a communication unit, a power transmission line image identification unit and a main control unit;
the communication unit acquires the power transmission line image shot by the camera and sends the power transmission line image to the main control unit;
the main control unit preprocesses the image of the power transmission line and then sends the image of the power transmission line to the image recognition unit of the power transmission line;
the power transmission line image recognition unit is internally provided with a depth convolution neural network image recognition model, and is used for extracting the characteristics of the power transmission line image and judging whether a fault occurs and the fault position.
Furthermore, the system also comprises an encryption unit which is used for carrying out remote communication encryption with the monitoring platform to realize data protection in the transmission process.
Further, the system also comprises a display unit for displaying fault alarm information and fault positions.
Further, when the image recognition unit of the power transmission line judges that a fault occurs, the fault type and the fault position are sent to the main control unit, and the main control unit sends the fault type and the fault position to the operation and maintenance monitoring platform of the power transmission line.
Further, the deep convolutional neural network image recognition model is obtained through training, and the training samples are selected from a sample set formed by preprocessing the power transmission line image, marking the fault type and the coordinate position.
Further, the deep convolutional neural network image recognition model utilizes a Darknet-53 network as a feature extraction network to extract deep-level features through a residual block, the deep-level features of different levels are respectively spliced with the shallow-level features after upsampling, and three feature maps with different sizes are generated to perform multi-scale recognition on targets existing in the sample picture.
The technical scheme of the invention has the following beneficial technical effects:
(1) compared with a characteristic extraction method based on a sliding window, the deep convolutional neural network based on characteristic map extraction is constructed, single-step classification and identification of the image are achieved, and the identification speed is improved.
(2) The invention adopts the deep convolutional neural network image recognition model deployed at the front end of the camera, improves the real-time performance of fault recognition, reduces the network transmission pressure and reduces the memory consumption of the server end.
Drawings
FIG. 1 is an overall flow chart of the present invention;
FIG. 2 is a flow chart of a picture sample preprocessing;
FIG. 3 is a structural diagram of a transmission line image recognition device based on a feature map deep convolutional neural network;
fig. 4 is a schematic structural diagram of a deep convolutional neural network model.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail with reference to the accompanying drawings in conjunction with the following detailed description. It should be understood that the description is intended to be exemplary only, and is not intended to limit the scope of the present invention. Moreover, in the following description, descriptions of well-known structures and techniques are omitted so as to not unnecessarily obscure the concepts of the present invention.
The invention discloses a power transmission line image identification method based on a deep convolutional neural network, which comprises the following steps:
(1) preprocessing the image of the power transmission line, marking the fault type and the coordinate position to form a sample set of model training;
(2) establishing a deep convolution neural network image recognition model based on feature map extraction, training and packaging the samples in the sample set to generate a model file;
(3) and deploying an image recognition device based on a deep convolutional neural network model at the front end of the camera to recognize the shot power transmission line image in real time.
The step (1) of preprocessing the power transmission line image and marking the fault type and the coordinate position comprises the following steps:
screening a large number of sample pictures generated by the power transmission line monitoring system by adopting a clustering analysis method, and selecting fault type image samples such as power transmission line external damage, mountain fire and the like;
the method comprises the steps that the Python programming language is utilized to realize unified size processing on all electric transmission line pictures, so that the electric transmission line pictures become standard sizes input by model training, and unified picture naming rules are adopted;
and (3) marking the positions of faults such as outburst faults, mountain fires and the like in the power transmission line image by using an image marking software tool to generate a sample marking file.
In the step (2), the step of establishing the deep convolutional neural network image recognition model based on feature map extraction is as follows:
constructing a deep convolutional neural network algorithm model based on feature map extraction by adopting a deep learning method of single-step classification identification;
taking the power transmission line image sample set and the labeled file set generated in the preprocessing in the step (1) as the input of a deep convolutional neural network model, setting the learning rate and the variation factor in a network model configuration file and the model training times, and performing model training until the precision meets the requirement or the training times are finished;
and selecting a test sample from the samples in the sample set, and testing the model identification accuracy precision of the sample. Then, adjusting the learning rate, the variation factor and the model training times, retraining the model, testing the model identification accuracy, comparing the model accuracy of the two times, and selecting the model file with higher identification accuracy to achieve the purpose of model optimization. .
And inputting the test sample into the deep convolution neural network image recognition model, verifying the model training effect, optimizing the network model according to the test result, and generating a model file.
The deep convolutional neural network image recognition model is shown in fig. 4, and comprises a Darknet-53 network as a feature extraction network, three feature maps with different sizes are generated to perform multi-scale recognition on targets existing in a sample picture, so that the recognition accuracy of the targets with different sizes is increased, and the recognition speed is high. And then screening out the target detection frame with the highest confidence coefficient by a non-maximum value inhibition method to achieve accurate identification of the target category and the frame.
Inputting 3-channel RGB format picture data, wherein the size is 416 × 416; the Darknet-53 network is used as a feature extraction network; the Darknet-53 network firstly carries out convolution conv on an input picture, carries out standardization processing through a BN layer, adopts a Relu function to add a nonlinear factor, then carries out residual block resn processing, and replaces pooling by the residual block to enable the network structure to be deeper and deeply extract target characteristics. And (2) resn: n represents a number, res1, res2, …, res8, indicating how many residual units are contained in the residual block.
Splicing concat tensors, and enabling the network to be shallow through upsampling; and splicing the upsampling of the middle layer and the later layer. The operation of stitching expands the dimensionality of the tensor so that deep and shallow features are stitched to one another. DBL 5 continues the convolution process 5 times to output the extracted feature map.
And finally, generating 3 feature maps with different scales for target detection in the sample picture. c denotes a failure category.
The package includes:
constructing a software architecture based on a deep convolutional neural network image recognition model, and deploying an image recognition algorithm model based on an artificial intelligence chip platform; the integrated device for transmitting the shot picture of the camera and identifying the image fault of the artificial intelligent chip is formed.
And inputting the image of the power transmission line to be identified into an image identification device based on a deep convolutional neural network, and acquiring the result of image identification, including the fault type of the power transmission line and the coordinate information in the picture.
The invention provides a power transmission line image recognition device based on a deep convolutional neural network, which is combined with the figure 3 and comprises a communication unit, a power transmission line image recognition unit, an encryption unit, a display unit and a main control unit;
the communication unit comprises a 4G/5G interface, an Ethernet port and an I2C interface, USB interface. The 4G/5G interface is used for communicating with the power transmission line operation and maintenance monitoring platform; the Ethernet port is used for controlling the camera to collect the image of the power transmission line, receiving the image of the power transmission line shot by the camera and sending the image of the power transmission line to the main control unit; i is2The interface C is used for power management; the USB interface is used for local upgrading.
And the encryption unit is used for carrying out remote communication encryption with the monitoring platform to realize data protection in the transmission process.
And the display unit is used for displaying the fault alarm information and the fault position.
The power transmission line image recognition unit is internally provided with a depth convolution neural network image recognition model, and is used for extracting the characteristics of the power transmission line image and judging whether a fault occurs and the fault position.
The main control unit adjusts the size of the power transmission line image and then sends the power transmission line image to the power transmission line image recognition unit, receives the result fed back by the power transmission line image recognition unit, outputs alarm information if a fault occurs, sends the fault position of the alarm information to the display unit for displaying, and sends the fault position of the alarm information to the power transmission line operation and maintenance monitoring platform through the 4G/5G interface.
In a specific embodiment, a deep convolutional neural network image recognition model based on feature map extraction is constructed, and the implementation mode of generating a model file is as follows:
the video and the image of the power transmission line are shot by a camera erected on a tower and transmitted to a background server end to be generated. The image sample set is preprocessed, and the specific flow is shown in fig. 2. The fault types identified by the transmission line comprise: external damage faults of construction machinery and the like and natural disaster faults of mountain fire. And screening the video and image samples to select a sample set with the transmission line fault information. The resolution of the picture sample set is uniformly modified, the image resolution is usually 1920 × 1080 or 1080 × 720, and the uniform processing is 416 × 416, and the image is used as a standard input of a network model. Processing the picture samples according to a uniform naming rule, labeling the fault types and the coordinate positions in the pictures to generate a labeled file, and adjusting the labeled file to start from (1,1) when the fault type coordinates start from the upper left corner of the pictures, namely from (0,0), so as to avoid the situation that the samples cannot be correctly input into the network model.
Inputting the image sample set and the corresponding annotation file set into a feature-map-based deep convolutional neural network model for training, wherein the feature-map-based deep convolutional neural network model adopts a YOLO v3 network algorithm with high image recognition speed, and sets appropriate model precision parameters according to the characteristics of the image sample set, so that the situation that the precision parameters are set too small and fall into local optimization is avoided.
And inputting the test picture sample into the image recognition model according to the model file generated by the model training result to verify the model recognition speed and accuracy, adjusting relevant parameters of model training or optimizing a sample set according to the test result to retrain the model, generating the model file, and improving the fault recognition accuracy.
And inputting the video and the image shot by the camera as a picture to be identified into an image identification device based on the deep convolutional neural network, and acquiring the result of image identification, wherein the result comprises the fault type of the power transmission line and the coordinate information in the picture, and the fault type comprises dangerous fault information such as external damage, mountain fire and the like.
In summary, the invention relates to a method and a device for identifying an image of a power transmission line based on a deep convolutional neural network, and the method comprises the steps of preprocessing the image of the power transmission line as a sample set of model training, performing unified picture size processing on sample pictures of faults such as outburst faults, forest fires and the like, labeling fault types, and generating a labeled file set; secondly, training a sample picture set by adopting a deep convolution neural network method based on feature map extraction to generate a model file; and finally, deploying an image recognition device based on a deep convolutional neural network model at the front end of the camera to realize the rapid and accurate recognition of the transmission line fault. The power transmission line image identification method and device based on the feature map deep convolutional neural network can improve timeliness of image identification of multiple fault types and reduce consumption rate of network resources and server memory resources for image identification of images transmitted back to a background server.
It is to be understood that the above-described embodiments of the present invention are merely illustrative of or explaining the principles of the invention and are not to be construed as limiting the invention. Therefore, any modification, equivalent replacement, improvement and the like made without departing from the spirit and scope of the present invention should be included in the protection scope of the present invention. Further, it is intended that the appended claims cover all such variations and modifications as fall within the scope and boundaries of the appended claims or the equivalents of such scope and boundaries.
Claims (10)
1. The power transmission line image identification method based on the deep convolutional neural network is characterized by comprising the following steps of:
(1) preprocessing the image of the power transmission line, marking the fault type and the coordinate position to form a sample set of model training;
(2) establishing a deep convolution neural network image recognition model based on feature map extraction, and training and packaging by adopting the samples in the sample set;
(3) and collecting the image of the power transmission line, identifying by adopting the packaged deep convolutional neural network image identification model, and judging whether a fault occurs and the fault position.
2. The method for identifying the image of the power transmission line based on the deep convolutional neural network of claim 1, wherein the preprocessing the image of the power transmission line and the labeling of the fault type and the coordinate position in the step (1) comprise:
screening the power transmission line images generated by the power transmission line monitoring system by adopting a clustering analysis method, and selecting various fault type image samples;
carrying out size processing on all the electric transmission line pictures by utilizing a Python programming language to enable the electric transmission line pictures to become standard sizes input by model training, naming and marking fault types by adopting a uniform picture naming rule;
and marking the fault positions of various fault types in the power transmission line image by using an image marking software tool to generate a sample marking file.
3. The method according to claim 2, wherein the fault types include power transmission line construction machinery, a tower crane, a crane and a mountain fire.
4. The method for recognizing the image of the power transmission line based on the deep convolutional neural network as claimed in claim 1 or 2, wherein in the step (2), a deep convolutional neural network image recognition model based on feature map extraction is established, and the steps of training and encapsulating by using the samples in the sample set are as follows:
constructing a deep convolutional neural network algorithm model based on feature map extraction by adopting a deep learning method of single-step classification identification;
adopting the samples in the sample set as the input of a deep convolutional neural network model, setting the learning rate, the variation factor and the model training times in a network model configuration file, and carrying out deep convolutional neural network model training until the precision meets the requirement or the training times are finished;
selecting a test sample from the samples in the sample set, and identifying accuracy precision of a deep convolutional neural network model of the test sample;
adjusting the learning rate, the variation factor and the model training times, retraining the deep convolutional neural network model, and testing the recognition accuracy of the deep convolutional neural network model;
comparing the accuracy rates of the two deep convolutional neural network models, selecting the model file of the deep convolutional neural network model with higher identification accuracy rate, and realizing model optimization.
5. The utility model provides a transmission line image recognition device based on degree of depth convolution neural network which characterized in that includes: the system comprises a communication unit, a power transmission line image identification unit and a main control unit;
the communication unit acquires the power transmission line image shot by the camera and sends the power transmission line image to the main control unit;
the main control unit preprocesses the image of the power transmission line and then sends the image of the power transmission line to the image recognition unit of the power transmission line;
the power transmission line image recognition unit is internally provided with a depth convolution neural network image recognition model, and is used for extracting the characteristics of the power transmission line image and judging whether a fault occurs and the fault position.
6. The device for recognizing the image of the power transmission line based on the deep convolutional neural network as claimed in claim 5, further comprising an encryption unit for encrypting the image in remote communication with the monitoring platform to realize data protection in the transmission process.
7. The device for identifying the power transmission line image based on the deep convolutional neural network as claimed in claim 5 or 6, further comprising a display unit for displaying fault alarm information and fault positions.
8. The device for identifying the image of the power transmission line based on the deep convolutional neural network as claimed in claim 5 or 6, wherein when the image identification unit of the power transmission line judges that a fault occurs, the type and the position of the fault are sent to the main control unit, and the main control unit sends the type and the position of the fault to the operation and maintenance monitoring platform of the power transmission line.
9. The device for recognizing the image of the power transmission line based on the deep convolutional neural network as claimed in claim 5 or 6, wherein the deep convolutional neural network image recognition model is obtained through training, and the training sample is selected from a sample set formed by preprocessing the image of the power transmission line, and marking the fault type and the coordinate position.
10. The device for identifying the image of the power transmission line based on the deep convolutional neural network as claimed in claim 5 or 6, wherein the deep convolutional neural network image identification model utilizes a Darknet-53 network as a feature extraction network to extract deep features through a residual block, the deep features of different layers are respectively spliced with the shallow features after upsampling, and three feature maps with different sizes are generated to perform multi-scale identification on the target existing in the sample picture.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010989285.1A CN112232133A (en) | 2020-09-18 | 2020-09-18 | Power transmission line image identification method and device based on deep convolutional neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010989285.1A CN112232133A (en) | 2020-09-18 | 2020-09-18 | Power transmission line image identification method and device based on deep convolutional neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112232133A true CN112232133A (en) | 2021-01-15 |
Family
ID=74107741
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010989285.1A Pending CN112232133A (en) | 2020-09-18 | 2020-09-18 | Power transmission line image identification method and device based on deep convolutional neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112232133A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112884018A (en) * | 2021-01-28 | 2021-06-01 | 广东电网有限责任公司 | Power grid line fault recognition model training method and power grid line inspection method |
| CN113781388A (en) * | 2021-07-20 | 2021-12-10 | 许继集团有限公司 | Image enhancement-based power transmission line channel hidden danger image identification method and device |
| CN113781729A (en) * | 2021-07-23 | 2021-12-10 | 广东电网有限责任公司广州供电局 | Power transmission line external force damage monitoring method and device based on intelligent gateway |
| CN113898284A (en) * | 2021-09-29 | 2022-01-07 | 国网山东省电力公司临朐县供电公司 | Insulation ladder for line loss inspection and using method thereof |
| CN114612855A (en) * | 2022-02-11 | 2022-06-10 | 江苏濠汉信息技术有限公司 | Power line hazard source detection system and method fusing residual error and multi-scale network |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109344753A (en) * | 2018-09-21 | 2019-02-15 | 福州大学 | A kind of tiny fitting recognition methods of Aerial Images transmission line of electricity based on deep learning |
| CN110069975A (en) * | 2018-12-25 | 2019-07-30 | 国网河北省电力有限公司电力科学研究院 | A kind of conducting wire neural network based opens burst recognition methods and system |
| CN110245644A (en) * | 2019-06-22 | 2019-09-17 | 福州大学 | A kind of unmanned plane image transmission tower lodging knowledge method for distinguishing based on deep learning |
| CN110689531A (en) * | 2019-09-23 | 2020-01-14 | 云南电网有限责任公司电力科学研究院 | Automatic power transmission line machine inspection image defect identification method based on yolo |
-
2020
- 2020-09-18 CN CN202010989285.1A patent/CN112232133A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109344753A (en) * | 2018-09-21 | 2019-02-15 | 福州大学 | A kind of tiny fitting recognition methods of Aerial Images transmission line of electricity based on deep learning |
| CN110069975A (en) * | 2018-12-25 | 2019-07-30 | 国网河北省电力有限公司电力科学研究院 | A kind of conducting wire neural network based opens burst recognition methods and system |
| CN110245644A (en) * | 2019-06-22 | 2019-09-17 | 福州大学 | A kind of unmanned plane image transmission tower lodging knowledge method for distinguishing based on deep learning |
| CN110689531A (en) * | 2019-09-23 | 2020-01-14 | 云南电网有限责任公司电力科学研究院 | Automatic power transmission line machine inspection image defect identification method based on yolo |
Non-Patent Citations (1)
| Title |
|---|
| 刘祥龙等: "《飞桨PaddlePaddle深度学习实战》", 31 August 2020 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112884018A (en) * | 2021-01-28 | 2021-06-01 | 广东电网有限责任公司 | Power grid line fault recognition model training method and power grid line inspection method |
| CN113781388A (en) * | 2021-07-20 | 2021-12-10 | 许继集团有限公司 | Image enhancement-based power transmission line channel hidden danger image identification method and device |
| CN113781729A (en) * | 2021-07-23 | 2021-12-10 | 广东电网有限责任公司广州供电局 | Power transmission line external force damage monitoring method and device based on intelligent gateway |
| CN113898284A (en) * | 2021-09-29 | 2022-01-07 | 国网山东省电力公司临朐县供电公司 | Insulation ladder for line loss inspection and using method thereof |
| CN114612855A (en) * | 2022-02-11 | 2022-06-10 | 江苏濠汉信息技术有限公司 | Power line hazard source detection system and method fusing residual error and multi-scale network |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112232133A (en) | Power transmission line image identification method and device based on deep convolutional neural network | |
| CN107808139B (en) | Real-time monitoring threat analysis method and system based on deep learning | |
| CN109977921B (en) | Method for detecting hidden danger of power transmission line | |
| CN113515655B (en) | Fault identification method and device based on image classification | |
| CN112528979B (en) | Transformer substation inspection robot obstacle distinguishing method and system | |
| CN114399672A (en) | Railway wagon brake shoe fault detection method based on deep learning | |
| CN112580657B (en) | Self-learning character recognition method | |
| CN109815884A (en) | Unsafe driving behavioral value method and device based on deep learning | |
| CN113642474A (en) | Hazardous area personnel monitoring method based on YOLOV5 | |
| CN112541393A (en) | Transformer substation personnel detection method and device based on deep learning | |
| CN106056069A (en) | Unmanned aerial vehicle image analysis-based forest land resource asset evaluation method and evaluation system | |
| CN112149551A (en) | Safety helmet identification method based on embedded equipment and deep learning | |
| CN115082813A (en) | Detection method, unmanned aerial vehicle, detection system and medium | |
| CN113378723B (en) | Automatic safety recognition system for hidden danger of power transmission and transformation line based on depth residual error network | |
| CN114120086A (en) | Pavement disease recognition method, image processing model training method, device and electronic equipment | |
| CN116310979B (en) | Image identification method, risk management and control platform and method, and safety management and control platform | |
| CN118038021A (en) | Transformer substation operation site foreign matter intrusion detection method based on improvement yolov4 | |
| CN113361968B (en) | Power grid infrastructure worker safety risk assessment method based on artificial intelligence and big data | |
| CN116958627A (en) | Model determination method and related device | |
| Rong et al. | A joint faster RCNN and stereovision algorithm for vegetation encroachment detection in power line corridors | |
| CN113989886A (en) | Crew identity verification method based on face recognition | |
| CN113239994A (en) | Power grid defect detection method and device based on YOLOv4-tiny algorithm, storage medium and electronic equipment | |
| CN116579609B (en) | Illegal operation analysis method based on inspection process | |
| CN118164196B (en) | Method and system for monitoring health state of coal conveying belt based on machine vision | |
| CN117892092B (en) | Intelligent acquisition method and system for realizing mine data based on big data |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210115 |
|
| RJ01 | Rejection of invention patent application after publication |