Video Stabilization by Feature-Block Motion Estimation

This paper presents a real-time video stabilization method for adjusting unstable frames captured by a vibrating camera. The basic idea is to calculate the vibration degree for compensating the vibration area by block motion detection and analyzing vibration type. A feature block is selected to obtain the vibration vector for checking if the frame is vibrant, and then the missing area is compensated and the overlapping area is adjusted. Experimental results show that the proposed algorithm can effectively reduce the frame vibration in real-time for the general vibration situations of moderate degree and compensate more than 21 pixels.