Title :
A computationally efficient IDCT algorithm
Author :
Bae, Seong-Ok ; Min, Seung-Jai
Author_Institution :
Inf. Technol. Lab., LG Corp. Inst. of Technol., Seoul, South Korea
Abstract :
This paper describes a computationally efficient inverse discrete cosine transform (IDCT) algorithm using the sparsity of IDCT input coefficients. IDCT input has many zero-valued coefficients by nature. Computing only non-zero valued coefficients results in a significant reduction in IDCT computation. Our new row-column decomposition based algorithm is suitable for calculating only non-zero valued coefficients in IDCT input. The simulation results show that the new algorithm is faster than conventional algorithms based on row-column decomposition. The fast computation feature enabled the use of only one 1-D IDCT unit for 2-D transform and the reduction of power consumption.
Keywords :
VLSI; decoding; discrete cosine transforms; video coding; computationally efficient IDCT algorithm; inverse discrete cosine transform; nonzero valued coefficients; power consumption; row-column decomposition based algorithm; sparsity; zero-valued coefficients; Computer architecture; Costs; Decoding; Digital signal processing; Discrete cosine transforms; Hardware; Image processing; Information technology; Kernel; Very large scale integration;
Conference_Titel :
Circuits and Systems, 1997. Proceedings of the 40th Midwest Symposium on
Print_ISBN :
0-7803-3694-1
DOI :
10.1109/MWSCAS.1997.662238
