Study on a Fast Motion Estimation Algorithm of Video Stabilization System

To resolve the problems that the local motions reduce the global motion estimation accuracy and traditional Gary Projection Algorithm could not deal with rotation motion, a real-time electronic image stabilization motion estimation method based on Fast Sub-block Gray Projection Algorithm is proposed. In the method, each image is divided into a number of sub-blocks, and sub-blocks are sifted according to their gray contrasts. By removing sub-blocks whose gray contrasts are lower than the threshold, the calculation amount of gray projection is reduced and the motion estimation accuracy is improved. Then gray projections are done in each sub-block remained to get local motion vectors. Global motion vector of the current image is calculated on the basis of local motion vectors of sub-blocks and the Affine Motion Model. The experiments results show that the algorithm is more accurate and efficient than the traditional Gray Projection Algorithm. Keyword: Video Stabilization, Motion Estimation, Gray Projection Algorithm, Gray Contrast