Fast Local Motion Estimation and Robust Global Motion Decision for Digital Image Stabilization

In this paper, we present a motion estimation method based on Gray-coded bit-plane matching for digital image stabilization systems. We achieve motion estimation by extracting four Gray-coded bit-planes from the image. Our proposed motion estimation approach can be realized using only Boolean functions, which significantly reduces the computational complexity while maintaining the accuracy of motion vectors. In addition, the global motion vector in the literature is decided directly from the local motion vectors in a simple way, which may obtain an unreliable resulting motion vector. We aim to this problem by proposing a more reliable algorithm to make the whole system more robust.