Video stabilization using Speeded Up Robust Features

Video stabilization is one of the most important enhancement techniques used to remove undesired motion in a video. Combination of global camera motion estimation along with motion separation determines the undesired motion, which is to be compensated to produce a stable video sequence. In this paper a novel method for robust video stabilization is proposed which uses Speeded Up Robust Features(SURF) as stable feature points to be tracked between frames for global motion estimation. Different measures are taken to select the most appropriate feature point trajectories. A discrete kalman filter is used to smoothen so estimated motion vectors and the resultant stabilized video is obtained by compensating the unstable motion.