Context adaptive variable length decoding system and method

Abstract

A system and method to perform context-adaptive variable length decoding (CAVLC) of transform coefficient levels for block-based motion-compensated decoding of moving pictures, corresponding to transform coefficients. The system and method includes complexity-reduction improvements in the coefficient level decoding process, such as simplified and extended range of Lev-VLC tables. Specifically, the number of Lev-VLC tables is extended from 5 to 7 and only 1 escape code (28-bit escape code) is used for tables Lev-VLC1to Lev-VLC6. The system and method also includes a simplified and improved table selection process. The table selection for the first Coefficient_level after trailing ones depends on total number of non-zero coefficients and number of trailing ones which are local variables within the CAVLC module. The table selection process for subsequent Coefficient_level's has been re-designed in such a way that the same logic path can be used to select Lev-VLC table for the next coefficient regardless of block modes and quantization parameters.

Dependencies are inferred from the claim text.