DocumentCode :
954599
Title :
Skew-circular/circular correlation decomposition of prime-factor DCT
Author :
Hu, N.-C. ; Lin, K.-C.
Author_Institution :
Dept. of Electron. Eng., Nat. Taiwan Inst. of Technol., Taipei, Taiwan
Volume :
142
Issue :
4
fYear :
1995
fDate :
8/1/1995 12:00:00 AM
Firstpage :
241
Lastpage :
246
Abstract :
An algorithm to decompose the prime-factor DCT into skew-circular/circular correlation (SCC/CC) by coset decomposition is proposed. The simplest case is when the two factors are odd and relatively prime. In this case, the DCT output components are split into six subgroups. Each subgroup contains three short-length CC or SCC matrix-vector products, and the three products can further be merged into only one short-length SCC or CC matrix-vector product. The six subgroups are independent, thus parallel computing is feasible. By fast computation of the short-length SCC and CC, this algorithm reaches the same or less number of multiplications compared to other efficient prime-factor algorithms
Keywords :
correlation theory; digital arithmetic; discrete cosine transforms; matrix decomposition; matrix multiplication; signal processing; algorithm; circular correlation decomposition; coset decomposition; fast computation; image processing; matrix-vector products; output components; parallel computing; prime-factor DCT; skew-circular correlation decomposition; subgroups;
fLanguage :
English
Journal_Title :
Vision, Image and Signal Processing, IEE Proceedings -
Publisher :
iet
ISSN :
1350-245X
Type :
jour
DOI :
10.1049/ip-vis:19952092
Filename :
465230
Link To Document :
بازگشت