Research On Real-Time Enhancement Methods Of Digital Video

With the development of digital video products and service, digital video has produced great economy influence to the industries of computer, telecommunication, and image. But it is true that the contrast of the video is low and the noise exists in the video sequence all the time due to various reasons. So it is a meaningful work to research on enhancement of digital video. The thesis is focused on the technique of real-time enhancement methods of digital video, which is sponsored by the project "Real-Time Enhancement of Gray Images". Some research is processed as follow:In the area of video de-noising, we did our research on various intra-frame de-noising methods on spatial domain, frequency domain and wavelet domain. And we did some experiments on these methods. By taking advantages of lifting-scheme's low computation complexity, we solved the problem of real-time video de-noising by using traditional wavelet transform. According to the characteristics of digital video that there exist strong correlations between frames, we introduced inter-frame filter methods, including temporal average filter and adaptive recursive filter. At last, a video de-noising method combined intra-frame and inter-frame was proposed, and experiments showed that this method had the good property in terms of real-time processing and de-noising effectiveness.And we also did some research on contrast enhancement of digital video based on histogram and wavelet transform.At last, all the algorithms were integrated to a set of simulation software we developed by using Directshow.