基于红外影像辅助的无人机自主巡检规划方法设计  

Design of Autonomous Patrol Planning Method for UAV Based on Infrared Image

作　　者：赵彦杰[1] 张四维 徐涛 曹世鹏 余万金 ZHAO Yanjie;ZHANG Siwei;XU Tao;CAO Shipeng;YU Wanjin(Nanjing Power Supply Company,Nanjing 210000,China;Allcorehatress(Beijing)Technology Co.Ltd.,Beijing 100193,China)

机构地区：[1]国网南京供电公司,江苏南京210000 [2]众芯汉创(北京)科技有限公司,北京100193

出　　处：《微型电脑应用》2022年第3期98-101,共4页Microcomputer Applications

摘　　要：为了提高对电网线路的自主巡检能力,设计了基于红外影像辅助技术的无人机自主巡检规划方法。在对电网线路实施红外成像处理的基础上,在机器视觉环境下提取红外影像的边缘轮廓特征量和纹理信息。然后结合对红外影像的信息融合和边缘区域检测结果提取影像的像素类别,通过模糊聚合融合处理方法检测影像的主成分特征,从而完成对影像信息增强处理。基于此,建立自适应图谱特征分集模型,再利用结合相似度特征分解方法实现对无人机自主巡检的视觉跟踪,从而完成无人机自主巡检规划。仿真结果表明:应用该方法后,无人机自主巡检规划的成像能力较好,空间规划能力较强。证明其有效提高了配电网的自主巡检能力。In order to improve the autonomous inspection ability of power grid lines, an autonomous inspection planning method of UAV is designed based on infrared image assistance technology. Based on the infrared imaging processing of power lines, the edge contour feature and texture information of the infrared image are extracted in the machine vision environment. The pixel category of the image is extracted by combining the information fusion of the infrared image with the detection results of the edge area, and the principal component features of the image are detected by the fuzzy aggregation fusion processing method, so as to complete the image information enhancement processing. Based on this, an adaptive atlas feature diversity model is established, and the visual tracking of UAV autonomous inspection is realized by combining the similarity feature decomposition method, so as to complete the autonomous inspection planning of UAV. The simulation results show that after the application of this method, the imaging ability of UAV autonomous inspection planning is better, and the spatial planning ability is stronger, which proves that it can effectively improve the autonomous inspection ability of distribution network.

关 键 词：红外影像辅助 无人机 自主巡检 信息增强 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]