Hand held mobile video stabilization using differential motion estimation

Abstract

The hand held mobile cameras suffer from different undesired slow motions during the scene capturing time. It is required to stabilize the video sequence by removing the undesired motion between the successive frames. Most of the existing methods are either very complex or does not perform well for slow and smooth motion of hand held mobile videos. In this paper a modified video stabilization algorithm for hand held camera videos is proposed which uses bicubic interpolation with Taylor series expansion to improve the estimation efficiency of the hierarchical differential global motion estimation. After motion estimation Gaussian kernel filtering is used to smoothen out estimated motion parameters. Then Inverse rotation smoothening is applied to remove the rotation effect from the stabilized transform chain. This reduces the accumulation error and minimizes missing image area significantly. The performance of the proposed algorithm is tested on various real time videos and also compared with existing algorithm. © 2012 Springer India Pvt. Ltd.