一种红外图像增强算法在无人机巡检输电线路上的应用  

Application of an infrared image enhancement algorithm in UAV inspection of transmission line

作　　者：陈科羽 石书山 CHEN Ke-yu;SHI Shu-shan(Guizhou Power Grid Co.,Ltd.Transmission Operation and Maintenance Branch,Guiyang 550002,China)

机构地区：[1]贵州电网有限责任公司输电运行检修分公司,贵州贵阳550002

出　　处：《电子设计工程》2020年第16期174-178,共5页Electronic Design Engineering

基　　金：中国科学院自动化研究所项目(61802408)。

摘　　要：由于输电线路巡检存在工作效率低,预算成本高、图像增强后对比度较低等问题,提出一种红外图像增强算法在无人机巡检输电线路上的应用研究。首先对输电线路红外图像增强前的预处理,包括图像去噪、非均匀性校正、边缘检测等。然后利用基于超像素分割的算法进行输电线路红外图像增强;后通过对比度、平均梯度、信息熵三项评价标准,对本文图像增强算法和以前图像增强算法对比分析。实验结果表明:本文算法进行红外图像增强后,对比度、平均梯度、信息熵三项标准值均高于其他算法,证明本算法的增强效果更好,改善了图像质量,方便了输电线路巡检工作。Due to the problems of low efficiency,high budget cost and low contrast after image enhancement,an infrared image enhancement algorithm is proposed for UAV inspection of transmission lines.Firstly,the preprocessing of transmission line infrared image before enhancement includes image denoising,nonuniformity correction,edge detection and so on.Then,the algorithm based on super-pixel segmentation is used to enhance the infrared image of transmission line.Then,the image enhancement algorithm in this paper and the image enhancement algorithm in Previous literature are compared and analyzed by three evaluation standards of contrast,average gradient and information entropy.The experimental results show that after infrared image enhancement,the contrast,average gradient and information entropy of the algorithm are higher than those of other algorithms,which proves that the enhancement effect of the algorithm is better,the image quality is improved,and the inspection work of transmission line is convenient.

关 键 词：红外图像增强算法 超像素分割 无人机检测 非均匀性校正 

分 类 号：TN62.59[电子电信—电路与系统]