DCT Domain Post-Processing Based on POCS

DCT 영역에서의 POCS에 근거한 후처리

  • 임창훈 (건국대학교 정보통신대학 인터넷미디어공학부)
  • Published : 2005.03.01

Abstract

Even though post-processing methods based on projections onto convex sets (POCS) have shown good performance for blocking artifact reduction, it is infeasible to implement POCS for real-time practical applications. This paper proposes DCT domain post-processing method based on POCS. The proposed method shows very similar performance compared to the conventional POCS method, while it reduces tremendously the computational complexity. DCT domain POCS performs the lowpass filtering in the DCT domain, and it removes the inverse DCT and forward DCT modules. Through the investigation of lowpass filtering in the iterative POCS method, we define kth order lowpass filtering which is equivalent to the lowpass filtering in the kth iteration, and the corresponding kth order DCT domain POCS. Simulation results show that the kth order DCT domain POCS without iteration gives very similar performance compared to the conventional POCS with k iterations, while it requires much less computations. Hence the proposed DCT domain POCS method can be used efficiently in the practical post-processing applications with real-time constraints.

POCS에 근거한 후처리 방식은 블록 결함 제거에 좋은 결과를 보여 주었지만, POCS는 반복적인 과정에 많은 연산을 요구하여 실시간 제약이 있는 실제적인 후처리를 위해서는 구현되기가 어렵다. 본 논문에서는 반복적인 기존의 POCS와 거의 비슷한 성능을 보이면서 계산 양을 획기적으로 감축한 DCT 영역에서의 POCS 방식을 제안한다. DCT 영역에서의 POCS 방식에서는 저역 통과 필터링을 DCT 영역에서 수행하여 역방향 DCT와 순방향 DCT 모듈을 제거하였다. 반복적인 POCS에서의 저역 통과 필터링을 분석하여, k번째 반복 과정에서의 필터링과 동일한 효과를 갖는 k차 저역 통과 필터링을 정의하고, 여기에 해당하는 k차 DCT 영역 POCS를 정의한다. 시뮬레이션 결과는 반복적 과정이 없는 k차 DCT 영역 POCS는 k번 반복적 과정을 거친 기존의 POCS 방식에 비하여 PSNR과 주관적 화질 면에서 거의 비슷한 성능을 보여 주면서, 훨씬 작은 계산 양을 필요로 한다. 본 논문에서 제안하는 DCT 영역에서의 POCS 방식은 실제적인 실시간 동작을 필요로 하는 후처리 구현을 위해서 효과적으로 사용될 수 있다.

Keywords

References

  1. S. -F. Chang and D. G. Messerschmitt, 'Manipulation and compositing of MC-DCT compressed video,' IEEE Journal of Selected Areas in Comm., vol.13, no.1, pp.1-11, Jan. 1995 https://doi.org/10.1109/49.363151
  2. N. Merhav and R. Kresch, 'Approximate convolution using DCT coefficient multipliers,' IEEE Trans. Circuits Syst. Video Technol., vol. 8, no. 1, pp.468-476, Aug. 1998
  3. C. Yim and M. A. Isnardi, 'An efficient method for DCT -domain image resizing with mixed field/frame mode macroblocks,' IEEE Trans. Circuits Syst. Video Technol., vol. 9, no. 5, pp. 696-700, Aug. 1999 https://doi.org/10.1109/76.780359
  4. H. Paek, R.-C. Kim, and S. U. Lee, 'On the POCS-based postprocessing technique to reduce the blocking artifacts in transform coded images,' IEEE Trans. Circuits Syst. Video Technol., vol. 8, no. 3, pp. 358-367, June 1998 https://doi.org/10.1109/76.678636
  5. S. H. Park and D. S. Kim, 'Theory of projection onto narrow quantization constraint set and its applications,' IEEE Image Processing, vol. 8, no. 10, pp. 1361-1373, Oct. 1999 https://doi.org/10.1109/83.791962
  6. D. Pearson and M. Whybray, 'Transform coding of images using interleaved blocks,' Proc. Inst. Elect. Eng., vol. 131, pp. 466-472, Aug. 1984
  7. H. S. Malvar and D. H. Staelin, 'The LOT: Transform coding without blocking artifacts,' IEEE Trans. Acoust., Speech, Signal Processing, vol. 37, pp. 553-559, Apr. 1989 https://doi.org/10.1109/29.17536
  8. B. Ramamurthi and A. Gersho, 'Nonlinear space-variant postprocessing of block coded images,' IEEE Trans. Acoust., Speech, Signal Processing, vol. 34, pp. 1258-1267, Oct. 1986 https://doi.org/10.1109/TASSP.1986.1164961
  9. A. Zakhor, 'Iterative procedure for reduction of blocking effects in transform image coding,' IEEE Trans. Circuits Syst. Video Technol., vol. 2, no. 1, pp. 91-95, Mar. 1992 https://doi.org/10.1109/76.134377
  10. Y. Yang, N. P. Galatanos, and A. K. Katsaggelos, 'Regularized reconstruction to reduce blocking artifacts of block discrete cosine transform compressed image,' IEEE Trans. Circuits Syst. Video Technol., vol. 3, no. 6, pp. 421-432, Dec. 1993 https://doi.org/10.1109/76.260198
  11. Y. Yang, N. P. Galatsanos, and A. K. Katsaggelos, 'Projection-based spatially adaptive reconstruction of block-transform compressed images,' IEEE Image Processing, vol. 4, no. 7, pp. 896-908, July 1995 https://doi.org/10.1109/83.392332
  12. Y. Jeong, I. Kim, and H. Kang, 'A practical projection-based postprocessing of block-coded images with fast convergence rate,' IEEE Trans. Circuits and Syst. Video Technol., vol. 10, no. 4, pp. 617-623, June 2000 https://doi.org/10.1109/76.845007
  13. A. Madisetti and A. N. Wilson, Jr., 'A 100 MHz 2-D DCT/IDCT processor for HDTV applications,' IEEE Trans. Circuits Syst. Video Technol., vol. 5, no. 2, pp. 158-165, 1995 https://doi.org/10.1109/76.388064
  14. N. I. Cho and S. U. Lee, 'Fast algorithm and implementation of 2-D discrete cosine transform,' IEEE Trans. Circuits and Syst., vol. 38, no. 3, pp. 297-305, Mar. 1991 https://doi.org/10.1109/31.101322
  15. R. Kresch and N. Merhav, 'Fast DCT domain filtering using the DCT and the DST,' IEEE Trans. Image Processing, vol. 8, no. 6, pp. 378-384, June. 1999
  16. 임창훈, 'DCT 영역에서의 2차원 분리 대청 선형 필터링,' 한국통신학회 논문지, 28권, 5C호, pp.461-467, 5월, 2003년