Abstract
HDTV signals have large information. A higher efficient HDTV coding technique is desired to enrich services of digital broadcasting. In the MPEG-2 video standard, overheads, such as motion vectors, prediction types and several control signals, are transmitted, and its ratio becomes larger at lower bitrates. Picture quality degradation is sometimes caused by this reason. In order to improve the picture quality by means to reduce the overheads, frame and field adaptive picture structures are proposed and motion vectors and prediction types are estimated in consideration for the amount of transmitted bits of the DCT coefficients in each macro-block. This paper presents those compression techniques and the effects.