基于SURF算法的输电线全景图像的特征提取

高压输电线跨距远、线路复杂,在充分分析电力线图像的特殊线性特性的基础上,文章提出了一种采用SURF算法完成输电线全景拼接并利用相位一致性提取线路特征的方法。首先采用SURF算法对部分电力线图片进行配准,并用RANSAC算法剔除错误的特征点对,然后拼接得到输电线全景图,对全景图采用相位一致性方法进行特征检测,最后提取经过标记的完整的单根电力线。对现场拍摄的部分输电线路图像进行了实验,结果提取出了完整、精确的单根电力线,说明该方法能提高输电线路弧垂计算的精度。

Power Line Panoramic Image Feature Extraction Based on SURF Algorithm

High-voltage transmission lines are usually span far and complex,on the base of fully analyzing the special linear characteristics of power line images,the SURF algorithm-based method is proposed to complete power line panorama stitching and then the phase congruency-based method is applied for extracting power line features.First,SURF algorithm is used in the matching of partial power line images and eliminating wrong feature points with the RANSAC algorithm.Then power line panoramic image is obtained by stitching.After that,phase congruency method is applied in detecting the features of panoramic image.Finally,the marked single power line is extracted.Transmission line images taken on-site are experimented.The method extracts a complete and accurate single power line,which can improve the accuracy of the transmission line sag calculations.