A New RTL Design Approach for a DCT/IDCT-Based Image Compression Architecture using the mCBE Algorithm
DOI:
https://doi.org/10.5614/itbj.ict.2012.6.2.3Abstract
In the literature, several approaches of designing a DCT/IDCT-based image compression system have been proposed. In this paper, we present a new RTL design approach with as main focus developing a DCT/IDCT-based image compression architecture using a self-created algorithm. This algorithm can efficiently minimize the amount of shifter -adders to substitute multiplier s. We call this new algorithm the multiplication from Common Binary Expression (mCBE) Algorithm. Besides this algorithm, we propose alternative quantization numbers, which can be implemented simply as shifters in digital hardware. Mostly, these numbers can retain a good compressed-image quality compared to JPEG recommendations. These ideas lead to our design being small in circuit area, multiplierless, and low in complexity. The proposed 8-point 1D-DCT design has only six stages, while the 8-point 1D-IDCT design has only seven stages (one stage being defined as equal to the delay of one shifter or 2-input adder). By using the pipelining method, we can achieve a high-speed architecture with latency as a trade-off consideration. The design has been synthesized and can reach a speed of up to 1.41ns critical path delay (709.22MHz).
Downloads
References
Salomon, D., Data Compression: The Complete Reference, 3rd Edition, New York: Springer-Verlag, 2004.
Loeffler, C., Lightenberg, A. & Moschytz, G.S., Practical Fast 1-D DCT Algorithms with 11-Multiplications, Proceedings ICASSP, 1989.
Jeong, H., Kim , J. & Cho, W.-k., Low-Power Multiplierless DCT Architecture Using Image Data Correlation, IEEE Transactions on Consumer Electronicsm, 2004.
Ruiz, G.A., Michell, J.A. & Buron, A., High Throughput ParallelPipeline 2-D DCT/IDCT Processor Chip, Journal of VLSI Signal Processing, 45, pp. 161-175, 2006.
Heyne, B. & Gotze, J., A Low Power and High Quality Implementation of The Discrete Cosine Transformation, Advance in Radio Science, 2007.
Kim, B.-I. & Ziavras, S.G., Low-Power Multiplierless DCT for Image/Video Coders, IEEE 13th International Symposium on Consumer Electronics, 2009.
Subramanian, P. & Reddy, A.S.C., VLSI Implementation of Fully Pipelined Multiplierless 2D DCT/IDCT Architecture for JPEG,Proceedings ICSP, 2010.
LSI Design Contest Official Website, Okinawa, Japan, http://lsicontest.com/2011/shiyou_3e.html (1 December 2010).
Sung, T.Y., Shieh, Y.S. & Sun, M.J., A High-Throughput and MemoryEfficiency 2D DCT Architecture based on Cordic Rotation, The 23rd Workshop on Combinatorial Mathematics and Computation Theory.