DocumentCode :
3251245
Title :
The dynamic range acquisition of DCT and IDCT algorithms
Author :
Ouyang, Wanli ; Xiao, Chuangbai ; Ju, Wenqi ; Song, Dandan
Author_Institution :
Coll. of Comput. Sci. & Technol., Beijing Univ. of Technol., China
fYear :
2005
fDate :
7-10 Aug. 2005
Firstpage :
429
Abstract :
A matrix-number inequality theory is proposed in this paper. Utilizing this theory, the data dynamic range for a famous DCT and IDCT algorithm is further determined. Employing the conclusion drawn from this dynamic range determination process, the average mean square error for its IDCT algorithm is decreased from 0.029422 to 0.014714. The proposed matrix-number inequality theory is applicable to any fast transform.
Keywords :
discrete cosine transforms; matrix algebra; mean square error methods; DCT algorithms; IDCT algorithms; discrete cosine transforms; dynamic range acquisition; matrix-number inequality theory; mean square error; Discrete cosine transforms; Dynamic range; Fixed-point arithmetic; Linear matrix inequalities; Matrix decomposition; Mean square error methods; Signal processing algorithms; Transform coding; VLIW; Video compression;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Circuits and Systems, 2005. 48th Midwest Symposium on
Print_ISBN :
0-7803-9197-7
Type :
conf
DOI :
10.1109/MWSCAS.2005.1594129
Filename :
1594129
Link To Document :
بازگشت