Research Article
Defect Detection Method of Overhead Line Pins Based on Multi-Sensor Data Acquisition of UAV
@INPROCEEDINGS{10.1007/978-3-031-50571-3_5, author={Xiaokaiti Maihebubai}, title={Defect Detection Method of Overhead Line Pins Based on Multi-Sensor Data Acquisition of UAV}, proceedings={Multimedia Technology and Enhanced Learning. 5th EAI International Conference, ICMTEL 2023, Leicester, UK, April 28-29, 2023, Proceedings, Part I}, proceedings_a={ICMTEL}, year={2024}, month={2}, keywords={UAV Multi-Sensor Data Acquisition Pin Image Preprocessing Image Segmentation Defect Detection}, doi={10.1007/978-3-031-50571-3_5} }- Xiaokaiti Maihebubai
Year: 2024
Defect Detection Method of Overhead Line Pins Based on Multi-Sensor Data Acquisition of UAV
ICMTEL
Springer
DOI: 10.1007/978-3-031-50571-3_5
Abstract
Common defects such as loose and missing pins on power fittings seriously affect the stable operation of the power system. In order to find and eliminate defects in time, a method for detecting pin defects in overhead lines based on multi-sensor data acquisition by UAV is proposed. UAV multi-sensors are used to collect images of overhead line pins, and grayscale, denoising and enhancement processing are performed on the images. On this basis, a watershed based image segmentation method is proposed to eliminate background interference. Defect detection is realized based on Faster-RCNN network, features are extracted through hierarchical residual convolution module, RPN is used to delineate regions of interest, Fast-RCNN network is applied to realize category confirmation, and detection results are output. The experimental results show that under the application of the research method, the AP values of the five types of samples are all above 0.9, indicating that the research method has a high defect detection accuracy.
