AN ADAPTIVE MULTIALPHABET ARITHMETIC CODING FOR VIDEO COMPRESSION

In this paper, the hardware implementation issues far an adaptive mult ialphabet arithmetic coder are discussed. A simple weighted history mo del is proposed to encode the video dare, This model uses a weighted f inite buffer to model the cumulative density function of the arithmeti c coder. The performance of the weighted history model is evaluated to gether with several other well-known models. To access, search, and up date the cumulative frequencies corresponding to model symbols in real rime, we present a low complexity multibase cumulative occurrence arr ay structure that can offer the probability information Tor high-speed encoding and decoding, for the application in video compression, the multialphabet arithmetic coding with weighted history model can be a g ood choice as the variable length coding of the video symbols.