DocumentCode :
1775186
Title :
Means and results of efficiency analysis for data compression methods applied to typical multimedia data
Author :
Kozhemiakina, Nadejda ; Ponomarenko, Nikolay ; Lukin, Vladimir ; Egiazarian, Karen ; Astola, Jaakko
Author_Institution :
Dept. of Signal Reception, Transm. & Process., Nat. Aerosp. Univ., Kharkov, Ukraine
fYear :
2014
fDate :
14-17 Oct. 2014
Firstpage :
12
Lastpage :
14
Abstract :
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.
Keywords :
data compression; discrete cosine transforms; multimedia computing; text analysis; DCT; data compression methods; data transmission systems; discrete cosine transform; large alphabet; multimedia data; multimedia information compression; statistical characteristics; test data sequences; text lossless compression; Arrays; Discrete cosine transforms; Image coding; Multimedia communication; Redundancy; Transform coding; analytical models; data compression; large alphabets; lossless compression; test data compression;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Infocommunications Science and Technology, 2014 First International Scientific-Practical Conference Problems of
Conference_Location :
Kharkov
Print_ISBN :
978-1-4799-7342-2
Type :
conf
DOI :
10.1109/INFOCOMMST.2014.6992281
Filename :
6992281
Link To Document :
بازگشت