电测与仪表2025，Vol.62Issue(3)：183-189,7.DOI:10.19753/j.issn1001-1390.2025.03.022
基于无人机巡检的风机叶片表面缺陷检测技术
UAV-based inspection of wind turbine blade surface defects detection technology
摘要
Abstract
Under the background of"double carbon",it is particularly important to vigorously develop new energy.Wind power generation is an important clean energy,and the scale of wind power is also expanding in the field of new energy.With the increasing scale of wind turbines,the damage probability of blades is also increasing.Aiming at the problems of high cost and poor working environment of large-scale wind turbine blade defect detection,a wind turbine blade surface defect detection method based on UAV image acquisition and digital image processing is proposed in this paper.According to the characteristics of images collected by UAV,this paper adopts the weighted average meth-od to realize gray processing,and then,the median filtering is applied to realize image noise reduction;the image enhancement is realized by CLAHE algorithm,which makes the details of target area and defect more clear and complete,and improves the detection efficiency.The feature information of defect is separated and extracted through image foreground segmentation and threshold processing,and the connected domain is framed to realize the detection of blade surface.The accuracy and error rate of defect images is calculated and tested by introducing per-formance evaluation index MIoU.The experimental results show that the detection accuracy of the proposed method for typical blade defects such as trachoma,scratch and crack is above 90％,especially the detection accuracy of crack defects can reach 95％,which verifies the effectiveness and accuracy of the algorithm in blade detection.关键词
风电叶片/叶片缺陷/对比度自适应直方图均衡化/缺陷检测/无人机Key words
wind turbine blade/blade defect/CLAHE/defect detection/unmanned aerial vehicle分类
动力与电气工程引用本文复制引用
谭兴国,张高明..基于无人机巡检的风机叶片表面缺陷检测技术[J].电测与仪表,2025,62(3):183-189,7.基金项目
新疆自治区自然科学基金(2022D01F46) （2022D01F46）
哈密科技项目(hmkj202107) （hmkj202107）
新疆自治区人才发展专项(202102) （202102）
