Transmission Line Galloping Detection Based on SURF Feature and FLANN Matching Algorithm

This paper first analyzes the causes of the formation of transmission line galloping, and then introduces the advantages and disadvantages of common online monitoring methods. Aiming at the problems of traditional image matching algorithms, such as fewer feature extraction points, high mismatch rate and slow matching speed, the solution is proposed. This paper uses the combination of SURF and FLANN matching algorithms to solve. The SUFR feature is quite ideal in terms of detail compared with the SIFT feature, and the calculation of the extremum using the Hessian matrix improves the speed of feature extraction, making it simple and efficient. FLANN algorithm adopts tree structure to realize storage and search, which can effectively solve the problem of slow matching of high-dimensional features. The experimental results show that the proposed method can meet the requirements of on-site real-time detection in the online dance monitoring of transmission lines, and has certain anti-interference ability for rotation and illumination changes.