测控技术2026，Vol.45Issue(9)：44-52,9.DOI:10.19708/j.ckjs.2026.08.226
基于立体视觉的输电线交叉跨越距离测量方法
Measurement Method for Crossing Distance of Transmission Lines Based on Stereo Vision
摘要
Abstract
In the measurement of the crossing distance of transmission lines,the inaccurate stereo vision model is caused by the lack of calibration of the left and right cameras of the stereo vision ranging device,and the lack of correction of camera distortion parameters leads to spatial node fitting deviation,resulting in ineffective judgment of camera pose and measurement errors.Based on this,a precise measurement method for the crossing distance of transmission lines using stereo vision is proposed.The measurement accuracy is improved by cali-brating the left and right cameras,constructing a stereo vision model,and incorporating camera distortion pa-rameter correction.Based on the calibration results,a stereo vision measurement model is constructed,and a spherical model is used to fit the spatial nodes of the three-dimensional scene of the transmission line,thereby completing coordinate alignment and reference position determination,and solving the problem of camera pose determination.By using the stereo matching method to set the energy cost function,the disparity map is ob-tained,and by combining with the pinhole imaging principle,accurate measurement of crossing distance is a-chieved.The experimental results show that this method can accurately measure the crossing distance of trans-mission lines under adverse weather conditions such as rain,snow and haze as well as in complex terrain,with a measurement deviation of less than 2 mm.It verifies that this method can reduce manual operation risks and improve power grid operation and maintenance efficiency.关键词
立体视觉/输电线/交叉跨越距离/测量方法/畸变参数/视图立体匹配Key words
stereo vision/transmission lines/crossing distance/measurement method/distortion parameters/stereo matching分类
信息技术与安全科学引用本文复制引用
沈海军,吴将,陈聪,杨德栋,郑忻,陈桓..基于立体视觉的输电线交叉跨越距离测量方法[J].测控技术,2026,45(9):44-52,9.基金项目
国网浙江省电力有限公司科技项目(5211JJ240002) （5211JJ240002）
