Huffman decoding using cascaded sub-table lookup method

摘要

A bitstream is sampled at an interval predetermined as a function of available processing power and memory storage capacity. The sampled bits are evaluated against a first layer Huffman decoding subtable. The subtable obtains the address of a second layer subtable from an indexed location identified by the sampled bits. A second bit sample is parsed to the designated second layer subtable, which outputs either an address for a third layer subtable, or a symbol and an associated code length.