Video quality plays an important role in different video-related services, because viewer experience is greatly affected by it. Information technology has evolved a lot over the last decades, and due to this former analog video has been replaced by modern digital video solutions. This changeover raised new problems, because video contents need orders of magnitude more space to store, than other contents like text, still image or audio. Professionals had to develop compression algorithms optimized specifically for video in order to remain storage and network transmission economical.
In my thesis I summarize the characteristics of human vision and history of development of digital video compression. Then I focus on a special branch, which skips the usual motion estimation and compensation method. Instead, it applies 3D DCT (where the third dimension corresponds to time). I examine the benefits and disadvantages of this solution.
Finally I design and implement a video compression algorithm based on 3D DCT, which will be considered under various aspects.