基于改进SGBM算法的输电线路覆冰厚度测量  

Measurement of Ice Thickness of Transmission Line Based on Improved SGBM Algorithm

作　　者：毛先胤 邢懿 罗国强 黄欢 董龙飞 鲁彩江 MAO Xianyin;XING Yi;LUO Guoqiang;HUANG Huan;DONG Longfei;LU Caijiang(Guizhou Electric Power Test Research Institute,Guiyang 550002,China;School of Mechanical Engineering,Southwest Jiaotong University,610031,China)

机构地区：[1]贵州电网有限责任公司电力科学研究院,贵阳550002 [2]西南交通大学机械工程学院机电测控系,成都610036

出　　处：《自动化与仪器仪表》2021年第11期23-26,31,共5页Automation & Instrumentation

基　　金：中国南方电网有限责任公司科技项目,输配电线路覆冰监测有效性提升技术研究(子课题3),编号:GZKJXM20191287。

摘　　要：传统的单目视觉对高压输电线路覆冰厚度的测量精度不高。为了提高覆冰厚度的准确性,提出了一种基于SGBM(Semi-Global Matching)的改进立体匹配算法。将Census变换与SGBM算法结合,从而提高了覆冰的匹配精度,进而提升覆冰厚度的测量精度。利用双目相机对模拟覆冰进行实验,实验结果验证了该算法的可行性和准确性,表明改进的算法能够显著地提高覆冰厚度的测量精度。Traditional monocular vision is difficult to accurately measure the thickness of ice on transmission lines.In order to improve the accuracy of measurement,an improved stereo matching algorithm based on SGBM(Semi-Global Matching)is proposed.The Census transform is combined with the SGBM algorithm to improve the matching accuracy and time.Experiments were performed on simulated ice coating with binocular cameras.The experimental results verified the feasibility and accuracy of the algorithm,indicating that the improved algorithm can significantly improve the measurement accuracy of ice coating thickness,and the matching time is significantly reduced.

关 键 词：双目视觉 输电线路 覆冰厚度 SGBM 立体匹配 

分 类 号：TP29[自动化与计算机技术—检测技术与自动化装置]