Power Line Feature Extraction Based on Optimized SURF Algorithm

In order to ensure the security and stability of the line, it is important that the sag of transmission line is regularly tested, and the key for the method of sag calculation based on the image processing is how to precisely extract the interested single power line feature. On the base of fully analyzing the special linear characteristics of power line images, the optimized SURF algorithm is proposed to detect power line image feature. By studying the spatial relationship of the extracted key points with SURF algorithm it is found that the key points which can best embody the characteristics of the linear image are in the same scale. But numerous and dense key points are extracted by SURF algorithm in texture-rich region, which is not conducive to the power line feature extraction, so SURF algorithm is optimized. Set the threshold of scale to retain the key points which is useful for extracting power line parameters, and by using the described information of the key points the power line parameters can be extracted accurately. Transmission line images are taken on the experiment; the results show that the optimization of the SURF algorithm can extract the sub pixel-level precision power line parameters, which can improve the accuracy of transmission line sag calculation.