Means and results of efficiency analysis for data compression methods applied to typical multimedia data
Nadejda Kozhemiakina, Nikolay N. Ponomarenko, Владимир Васильевич Лукин, Karen Egiazarian, Jaakko T. Astola · 2014
A problem of lossless compression of a text with a large alphabet, typical for compressing multimedia information in data transmission systems, is considered. A method of forming test data sequences for a given alphabet size is proposed. It allows calculating theoretically achievable compression ratio for the generated test sequences. It is shown that formed test sequences have statistical characteristics similar to the quantized coefficients of the Discrete Cosine Transform (DCT) of blocks in images or video frames. Six test sequences are generated (for alphabets sizes 2, 4, 8, 16, 32 and 64 bytes). It is shown that existing archivers do not always provide high enough efficiency for compressing these sequences and their efficiency reduces with an alphabet size increase.