An adaptive implementation of the MPEG encoder

Abstract

Due to the huge amount of data contained in digital video signals, the compression of video signals becomes crucial and important when the storage and transmission of digital video information are required. Most compression algorithms use motion compensated predictive coding to remove temporal redundancy and transformation coding to remove spatial redundancy. Additional entropy coding algorithms are also used to further reduce the amount of data in representing the compressed video information. With such a structure, a MPEG standard for video compression has been established by the International Standard Organization (ISO). The same compression techniques can be easily generalized to encode high definition (HD) video. In the paper, the authors propose a novel structure for encoding HD video, which provides an adaptive coding structure without any additional computations than those conventionally used. Since the temporal correlation is utilized during the encoding process to remove temporal redundancy, it is expected to produce a savings of bit rate in the reconstructed HD video.<>