A higher efficient HDTV coding technique is desired to enrich digital broadcasting services. In the MPEG-2 video coding, which is commonly used for digital broadcasting, a motion vector rate is almost constant and not related to the coding bit rate. Therefore, at a lower bit rate, the motion vector rate is generally high and the quantiser scale becomes large to reduce the DCT coefficient rate. This is the major factor that causes remarkable picture degradation by HDTV low bit rate coding. To improve picture quality at the lower bit rate, the motion vector rate should be suppressed. This paper describes an improvement method of picture quality by selecting the coding mode and the motion vectors so that give the minimum amount of the transmitted bits in a macroblock.
