Real-Time Automatic Route Generation for Unmanned Aerial Vehicle Based Patrol Inspection in Power Distribution System

Wang, Jianzhong; Su, Yanyan; Shen, Lang; Li, Kai

doi:10.1109/iaeac54830.2022.9929755

Cited by 3 publications

(1 citation statement)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In comparison, with the help of drones, remote sensing equipment for power inspection [6] [7], although to avoid the aforementioned methods exist interference problems, and relatively higher safety factor, according to whether the use of deep learning algorithms can be divided into traditional image processing detection algorithm and deep learning detection algorithm. The traditional image processing algorithm, mainly through such ways as: edge detection, image morphology and wavelet transform detection, such algorithms for image quality requirements are high and the generalization of the model is weak, therefore, difficult to be competent for complex scenes under the task of power inspection; and deep learning algorithms have stronger generalization, so can be better suited to complex scenes under the task of power inspection.…”

Section: Introductionmentioning

confidence: 99%

Real-time power line segmentation detection based on multi-attention with strong semantic feature extractor

Zhao,

Ji,

Liang

et al. 2023

J Real-Time Image Proc

View full text Add to dashboard Cite

This paper proposes a multi-objective power inspection method based on multiple attention mechanisms and strong semantic feature extraction. Based on the DeepabV3+ network model, the CBAM attention mechanism is introduced into the MobileNetV2 network to enhance the contextual information interaction capability; the ASPPAttention fast feature fusion structure is proposed to achieve fast extraction of multi-dimensional valid information by designing deep separable convolutions with different perceptual fields, and to enhance the pixel-level feature encoding capability with the CA attention mechanism; the lightweight feature encoding capability is proposed. The proposed lightweight inverted convolutional decoder structure improves the feature extraction capability of the model by designing an inverted bottleneck convolutional structure in two quadruple downsampling layers with a low number of parameters, and introduces the CA attention mechanism to avoid the heterogeneous gap; in the training process, the convergence speed of the model is accelerated by using migration learning, and Dice Loss is introduced to avoid the effect of the number of samples on the generalization of the

show abstract

Section: Introductionmentioning

confidence: 99%

Real-time power line segmentation detection based on multi-attention with strong semantic feature extractor

Zhao,

Ji,

Liang

et al. 2023

J Real-Time Image Proc

View full text Add to dashboard Cite

show abstract

An Inspection Mode Based on Unmanned Aerial Vehicle for Photovoltaic System

Xiao¹,

Zhou²,

Xu³

et al. 2022

2022 IEEE 5th Advanced Information Management, Communicates, Electronic and Automation Control Conference (IMCEC)

View full text Add to dashboard Cite

Real-time power inspection method based on multi-attention and strong semantic feature extractor

Zhao¹,

Ji²,

Liang³

et al. 2023

Preprint

View full text Add to dashboard Cite

This paper proposed a multi-objective power inspection method based on multiple attention mechanisms and strong semantic feature extraction.Based on the Deeplabv3+ network model,the CBAM attention mechanism is introduced into the MobileNetV2 network to enchance the contextual information interaction capability; the ASPPAttention fast feature fusion structure is proposed to achievefast extraction of multi-dimensional valid information by designing deep separa-ble convolutions with different perceptual fields, and to enhance the pixel-levelfeature encoding capability with the CA attention mechanism; the lightweightfeature encoding capability is proposed. The proposed lightweight inverted con-volutional decoder structure improves the feature extraction capability of themodel by designing an inverted bottleneck convolutional structure in two quadru-ple downsampling layers with a low number of parameters, and introduces the CAattention mechanism to avoid the heterogeneous gap; in the training process, theconvergence speed of the model is accelerated by using migration learning, andDice Loss is introduced to avoid the effect of the number of samples on the gener-alization of the model. The experimental results show that the MIoU of the powersegmentation inspection method based on attention and information decouplingcan reach 48.5%, the accuracy rate can reach 97.5%, and the detection speed ofthe model can reach 40.8 FPS, which is a better balance of speed and accuracycompared with the network models of HRNet, PSPNet and DeeplabV3+.

show abstract

Real-Time Automatic Route Generation for Unmanned Aerial Vehicle Based Patrol Inspection in Power Distribution System

Cited by 3 publications

References 11 publications

Real-time power line segmentation detection based on multi-attention with strong semantic feature extractor

Real-time power line segmentation detection based on multi-attention with strong semantic feature extractor

An Inspection Mode Based on Unmanned Aerial Vehicle for Photovoltaic System

Real-time power inspection method based on multi-attention and strong semantic feature extractor

Contact Info

Product

Resources

About