Recovery of blurred video signals using iterative image restoration combined with motion estimation

This paper proposes the recovery technique of blurred video signals using the new regularized constrained iterative image restoration (RCIIR) algorithm, which is combined with motion estimation. We enhanced the subjective video quality and improved the resolution of video signals by applying the new RCIIR algorithm, which had been developed with blurred still images, to the moving objects in the video sequences. Based on the extracted motion information of moving objects from the motion estimation technique, we could obtain the approximated point spread function, which was modelled as uniform motion blur. We then applied the RCIIR algorithm to the blurred moving objects so that we could improve the resolution and quality of the videos. The experimental results showed that the boundaries of blurred objects became clearer and the overall subjective quality of the restored video sequences was improved. Our proposed recovery technique was suitable to enhance degraded video signals due to motion blurs and it was applicable to non-real time video processing to enhance the subjective quality and resolution