Digital Image Stabilization Using Histogram-Based Sorting

In this paper, an efficient algorithm for the fast full search is proposed to reduce the computational cost of block matching algorithm for local motion estimation in digital image stabilization (DIS). This method based on histogram sorting. For the current block, the proposed fast matching method adaptively sorts the matching order pixel-by-pixel for each block. Also, partial distortion elimination (PDE) method was used to terminate the improper candidate blocks and reduce computation for block matching. Also a new method for extract global motion vector from local motion vectors of 4 areas is proposed. Some test sequences were simulated and the experimental results demonstrate that the proposed algorithm achieves less computational load.