Real-time video stabilization by adaptive fuzzy filtering

Digital video stabilization (DVS) allows acquiring video sequences without disturbing jerkiness, removing unwanted camera movements. In this paper, we propose a novel DVS algorithm that adaptively compensates the camera jitters utilizing an adaptive fuzzy filter on the global motion of video frames. The adaptive fuzzy filter is a simple infinite impulse response (IIR) filter which is tuned adaptively by a fuzzy system. The fuzzy system has two inputs which are used as quantitative representations of unwanted and intentional camera motion. The fuzzy system is also tuned adaptively during operation according to the characteristics of camera jitters. The global motion of video frames is estimated based on the block motion vectors which are resulted by video encoder during motion estimation operation. The proposed method also utilizes an adaptive criterion for validating of motion vectors. Experimental results have indicated a good performance for the proposed DVS algorithm.