100% 하드웨어 효율을 갖는 블록기반의 이차원 이산 웨이블렛 변환 필터 설계

Design of a Block-Based 2D Discrete Wavelet Transform Filter with 100% Hardware Efficiency

  • 김주영 (가톨릭대학교 정보통신 전자공학부) ;
  • 박태근 (가톨릭대학교 정보통신 전자공학부)
  • Kim, Ju-Young (Department of Information, Communication, and Electronic Engineering, The Catholic University of Korea) ;
  • Park, Tae-Guen (Department of Information, Communication, and Electronic Engineering, The Catholic University of Korea)
  • 투고 : 2010.07.21
  • 심사 : 2010.11.24
  • 발행 : 2010.12.25

초록

본 논문에서는 하드웨어 효율이 100%가 되는 2차원 이산 웨이블렛 변환 필터 구조를 제안한다. 전체 구조는 두 채널 QMF PR Lattice 필터로 구성된 1차원 DWT 필터 4개로 구성되었다. 1 레벨부터 J 레벨까지 순차적으로 수행함으로써 메모리 사용을 최소화 하면서도 하드웨어 효율이 100%가 되도록 설계하였으며 필터 입력 데이터를 구성해주는 DFC구조와 DCU구조를 제안하였다. 인접한 4개의 데이터를 동시에 입력 받아 처리함으로써 동시에 행방향과 열방향 DWT를 수행하므로 $N{\times}N$ 이미지를 처리하는데 $N^2(1-2^{-2J})/3$ 사이클이 소요되며 이 때 필요한 저장공간은 약 2MN-3N이다. 기존의 2D DWT 구조와 비교해 보았을 때 하드웨어 효율과 동작 속도가 향상되었으며 두 개의 1D DWT를 직렬로 연결하므로 임계경로를 감소시키기 위해서 최대 4 단까지 파이프라인을 적용하여 임계경로를 향상시킬 수 있다. 제안된 구조는 VerilogHDL로 모델링되고 동부아남 $0.18{\mu}m$ 표준셀로 합성되어 검증되었다.

This paper proposes a fully-utilized block-based 2D DWT architecture, which consists of four 1D DWT filters with two-channel QMF PR Lattice structure. For 100% hardware utilization, we propose a new method which processes four input values at the same time. On the contrary to the image-based 2D DWT which requires large memories, we propose a block-based 2D DWT so that we only need 2MN-3N of storages, where M and N stand for filter lengths and width of the image respectively. Furthermore, the proposed architecture processes in horizontal and vertical directions simultaneously so that it computes the DWT for an $N{\times}N$ image within a period of $N^2(1-2^{-2J})/3$. Compared to existing approaches, the proposed architecture shows 100% of hardware utilization and high throughput rate. However, the proposed architecture may suffer from the long critical path delay due to the cascaded lattices in 1D DWT filters. This problem can be mitigated by applying the pipeline technique with maximum four level. The proposed architecture has been designed with VerilogHDL and synthesized using DongbuAnam $0.18{\mu}m$ standard cell.

키워드

참고문헌

  1. Iain E. G. Richardson, "H.264 and MPEG-4 Video Compression," John Willey & Sons, 2003.
  2. S. Mallat, "A theory for multiresolution signal decomposition: The wavelet representation," IEEE Trans. Pattern Anal. and Machine Intell., vol.11, no.7, pp.674-693, 1989. https://doi.org/10.1109/34.192463
  3. T. Ryan, L. Sanders, H. Fisher, and A. Iverson, "Image compression by texture modeling in the wavelet domain," IEEE Trans. Image Process., vol.5, pp.26-36, 1996. https://doi.org/10.1109/83.481668
  4. A. Skodras, C. Christopoulos, and T. Ebrahimi, "JPEG2000: The Upcoming Still Image Compression Standard," Proceedings of 11th conf. of Pattern Recognition, pp.359-366, 2000.
  5. M. Vishwanath, R. Owens, and M. Irwin, "VLSI architectures for the discrete wavelet transform," IEEE Trans. CAS-II, vol.42, no.5, pp.305-316, 1995.
  6. T. Denk and K. Pahri, "VLSI architectures for Lattice structure based orthogonal discrete wavelet transform," IEEE Trans. CAS-II, vol.44, no.2, pp.129-132, 1997.
  7. A. Lewis and G. Knowles, "VLSI architecture for 2D Daubechies wavelet transform without multipliers," Electron. Lett., vol.27, no.2, pp.171-173, 1997.
  8. S. Paek and L. Kim, "2D DWT VLSI architecture for wavelet image processing," Electron.Lett., vol.34, no.6, pp.537-538, 1998. https://doi.org/10.1049/el:19980387
  9. J. Kim, Y. Lee, T. Isshiki, and H. Kunieda, "Scalable VLSI architecture for lattice structure-based discrete wavelet transform," IEEE Trans. CAS-II, vol.45, no.8, pp.1031-1043, 1998.
  10. C. Yu and S. Chen, "Design of an efficient VLSI architecture for 2D discrete wavelet transforms," IEEE Trans. Consumer Elect., vol.45, no.1, pp.135-140, 1999. https://doi.org/10.1109/30.754428
  11. F. Marino, "A Double-Face Bit-serial Architecture for the 1D Discrete Wavelet Transform," IEEE Trans. on Circuits & Systems II-Analog & Digital Signal Preocessing, vol.47, no.1, pp.65-71, 2000. https://doi.org/10.1109/82.818896
  12. T. Park and S. Jung, "High Speed Lattice Based VLSI Architecture of 2D Discrete Wavelet Transform for Real-Time video Signal Processing," IEEE Trans. on Consumer Electronics, vol.48, no.4, pp.1026-1032, 2002.
  13. P. Vaidyanathan, "Multirate systems and filter banks", Prentice-Hall, 1993.