Panoramic Monitoring Image Stitching Technology for Aerial Workers in Multi-View Mode
Yan Li, Peiyang Yan, Wu Li, Guang Liang, Guoqing Chen, Shuobing Chen · 2024
In the aerial operation scenario of transmission lines, there are a lot of horizontal and vertical movement of operators. In order to achieve omnibearing and no dead end tracking of the operators and improve the level of safety supervision and management of the operation sites, wide-angle lenses and zoom lenses are set in the operation sites to monitor operators. Image stitching technology is used for multi-view image fusion panoramic display. This paper proposes a SURF image feature extraction algorithm based on brightness correction, combined with FLANN and RANSAC image matching algorithms, to improve the accuracy of image stitching feature matching and the robustness of the system. Compared with SURF+RANSAC algorithm and ORB+ RANSAC algorithm, the matching success rate of the proposed algorithm is significantly improved, and the matching efficiency and feature point detection efficiency are both good. Therefore, the image stitching algorithm can solve the problem that the implementation of safety protection measures during the movement of workers who cannot be supervised in the aerial operation scenario, and has strong real-time performance, high accuracy and good robustness.