A Robust Image Sequence Registration Algorithm for Videosar Combining Surf with Inter-Frame Processing
Zihan Li, Zhen Dong, Anxi Yu, Zhihua He, Xiaoxiang Zhu · 2019
VideoSAR (video synthetic aperture radar) can provide continuous surveillance over a specific region, and shadow can be utilized to detect moving target. However, feature variations (unstable scattering properties and continuous motion) may seriously damage the precision of registration. Speeded Up Robust Feature (SURF) is one of the most popular method for feature detection and SAR image registration. So in this paper, a robust image sequence registration algorithm for videoSAR is proposed based on the combination of SURF and inter-frame processing. SURF is firstly introduced to videoSAR registration, whereas the continuous feature variation through the entire image sequence, showing a severe loss of robustness . Aiming at this defect, SURF and the idea of inter-frame processing are integrated and forming a novel algorithm. Simulation experiments underline the effectiveness and robustness of the proposed method in videoSAR image sequence registration.