Denoising Of Video Sequences

With the development of digital video products and service,digital video has produced great economy influence to the industries of computer,telecommunication,image. But noise exist in the video sequence all the time,which influence the visual effect and further processing on video. So the research of denoising on the video sequence is a meaningful work.After analyzing the problem of denoising systematically and deeply,we did our research work on Spatial domain,Frequency domain,Temporal domain,Color domain and Wavelet domain respectively to decompress the noise in the video sequence. And according to the characteristics of the noise,we provided some Algorithms:Wiener filter based on noise characteristic,multi-threshold wavelet filter,Temporal-Spatial-Frequency domain combined filter based on the motion estimation. Using these Algorithms,we can eliminate the noise in the video,and can reserve the useful details.The content of this paper is organized as following. Chapter one introduced the development and trend of the technology on video sequence denoising. Chapter two illustrate the theories and knowledge which are useful to decompress the noise in the video,and emphatically discussed the wavelet transform theory. Chapter three we introduced some pivotal algorithms and filters in theory,and on this basis,we designed several kinds of effective algorithms. Chapter four we realized the algorithms and analyzed these different results. Chapter five summarized the work that we have done,and expect the future of its development.