Video noise reduction based on motion complexity classification

Video sequences noise reduction are of great potential benefit to the video industry, as they promise the means to evaluate the performance of acquisition, display, coding and communication systems. Many researchers have focused on developing digital video sequences noise reduction methods. However, to be widely applicable a noise reduction must also work over a wide range of quality. In this paper, we have developed novel video sequence noise reduction method. According to motion complexity, video noise can be reduced with temporal and spatial noise reduction. They could quantify the spatial distortion and the type of the distortion. Experimental results show the validity of our approach and our quality metric performs better than the existing methods.