基于主元梯度直方图的输电线路障碍物检测  

Detection for transmission line obstacles based on principal component gradient histogram

作　　者：张峰 郭锐 程志勇 雍军 傅思遥[3] 韩立伟[3] 杨军[3] 贾乐刚 ZHANG Feng;GUO Rui;CHENG Zhiyong;YONG Jun;FU Siyao;HAN Liwei;YANG Jun;JIA Legang(Electric Power Research Institute of Shandong Power Supply Company of State Grid, Jinan 250000, China;Shandong Luneng Intelligent Technology Limited Corporation, Jinan 250000, China;School of Electrical Engineering, Wuhan University, Wuhan 430072, China)

机构地区：[1]国网山东省电力公司电力科学研究院,济南250000 [2]山东鲁能智能技术有限公司,济南250000 [3]武汉大学电气工程学院,武汉430072

出　　处：《计算机工程与应用》2016年第15期254-259,265,共7页Computer Engineering and Applications

摘　　要：基于视觉导航的输电线路巡检机器人在智能电网和输电线路巡检中有广泛的应用。提出一种基于主元方向梯度直方图特征的快速分类检测方法用于在线障碍物检测与识别。与传统基于几何结构基元的方法相比,该方法能在不提升计算复杂度的情况下显著提升识别精度,明显改善了识别算法在野外大范围复杂背景与光照影响下的性能。实验结果表明,该方法能达到精度和速度的性能平衡。Vision based powerline inspection robot navigation plays an essential role in modern smart grid, power system,and powerline inspection. This paper presents a fast and accurate online object classification approach using the combinationof Histogram of Oriented Gradient(HOG)features and Support Vector Machine(SVM). Compared with traditionalenvironmental constricted geometric based methods, the proposed approach significantly improves the classification performancewithout losing computing efficiency. In particular, the method works satisfactorily in large-scale complex backgroundand illumination variation scenario. Experiment results show that the proposed approach can achieve the balancebetween accuracy and efficiency.

关 键 词：输电线路检测 目标识别 方向梯度直方图 支持向量机 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]