River publishers
Yüklə 1,18 Mb.
|
Books on H.264 :1. I. E. G. Richardson, “H.264 and MPEG-4 video compression: Video coding for Next Generation Multimedia,” New York, Wiley, 2003. 2. S. Okubo et al, “H.264/AVC Textbook (in Japanese, title translated),” Tokyo, Impress Publisher, 2004. 3. S. Ono, T. Murakami and K. Asai, Ubiquitous technology, “Hybrid Video Coding – MPEG-4 and H.264 (In Japanese, title translated)”, Tokyo, Ohmsha Press, 2005. 4. L. Chiariglione (Editor), “The MPEG Representation of Digital Media”, Springer, 2012. H.264 Standard, JM SOFTWARE1. ITU – T and ISO/IEC JTC 1, “Advanced video coding for generic audiovisual services,” ITU-T Rec. H.264 & ISO/IEC 14496-10, Version 1, May 2003; version 2, Jan 2004; Version 3 (with High family of profiles), sept 2004; version 4, July 2005 [online] Available: http://www.itu.int/rec/T-REC-H.264. JM 19.0 Reference software (6/19/2015) also updates on KTA. DCT references DCT1N. Ahmed, M. Natarajan and K.R. Rao, “Discrete Cosine Transform,” IEEE Trans. Computers, vol. C-23, pp. 90–93, Jan. 1974. DCT2 W. Chen, C. H. Smith, and S. C. Fralick, “A fast computational algorithm for the discrete cosine transform,” IEEE Trans. Commun., vol. COM-25, pp. 1004–1009, Sept. 1977. DCT3 M. Vetterli and A. Ligtenberg “A discrete Fourier-cosine transform chip,” IEEE Journal on Selected Areas of Communications, vol. 4, pp. 49–61, Jan. 1986. DCT4 W.K. Cham and R.J. Clarke, “Application of the principle of dyadic symmetry to the generation of orthogonal transform,” IEE Proc. F: Communications, Radar and Signal Processing, vol. 133, no. 3, pp. 264–270, June 1986. DCT5 C. Loeffler, A. Lightenberg and G. Moschytz, “Practical fast 1-D DCT algorithms with 11 multiplications,” Proc. IEEE ICASSP, vol. 2, pp. 988-991, Feb. 1989. DCT6 W.K. Cham, “Development of integer cosine transforms by the principle of dyadic symmetry”, IEE Proc. Communications, Speech and Vision, vol. 136, issue 4, pp. 276-282, 1989. DCT7 M. Masera, M. Martina and G. Masera, “Adaptive approximated DCT architectures for HEVC”, IEEE Trans. CSVT (Early access).This has several references related to integer DCT architectures. Same as [E381] DCT8 S.-H. Bae, J. Kim and M. Kim, “HEVC-Based perceptually adaptive video coding using a DCT-Based local distortion detection probability”, IEEE Trans. on Image Processing, vol. 25, pp. 3343-3357, July 2016. Same as [E377]. DCT9 Z. Zhang et al, “Focus and blurriness measure using reorganized DCT coefficients for auto focus” IEEE Trans. CSVT, Early access. DCT10. S.-H. Bae and M. Kim, “DCT-QM; a dct-BASED QUALITY DEGRADATION METRIC FOR IMAGE QUALITY OPTIMIZATION PROBLEM”, IEEE Trans. IP, vol.25, pp. 4916-4930, Oct. 2016. DCT11 M. Jridi and P.K. Meher, “A scalable approximate DCT architectures for efficient HEVC compliant video coding” IEEE Trans. CSVT, (early access) See also P.K. Meher et al, “Generalized architecture for integer DCT of lengths N = 8, 16 and 32, IEEE Trans. CSVT, vol. 24, pp.168-178, Jan. 2014. DCT12 V. Dhandapani and S. Ramachandran, “Area and power efficient DCT architecture for image compression”, EURASIP Journal on Advances in Signal Processing, vol. , 20142014:180 DCT13 M. Narroschke, “Coding efficiency of the DCT and DST in hybrid video coding”, IEEE J. on selected topics in signal processing, vol. 7, pp.1062 – 1071, Dec. 2013. DCT14 G. Fracastoro, S.M. Fosson and E. Magli, “Steerable discrete cosine transform”, IEEE Trans. IP, vol. 26, pp. 303-314, Jan. 2017. Same as [E392]. DCT15 B. Zeng and J. Fu, “Directional discrete cosine transforms-a new framework for image coding,” IEEE Trans. CSVT, vol. 18, pp. 305-313, March 2008. DCT16 J.-S. Park et al, “2-D large inverse transform (16x16, 32X32) for HEVC (high efficiency video coding)”, J. of Semiconductor Technology and Science, vol.12, pp.203-211, June 2012. Same as [E364]. DCT17 J. Dong, et al, “2-D order-16 integer transforms for HD video coding,” IEEE Trans. CSVT, vol. 19, pp. 14621474, Oct. 2009.
DCT22 J. Lee at al, “A compressed-domain corner detection method for a DCT-based compressed image”, IEEE ICCE, Las Vegas, Jan. 2017. DCT23 Z. Qian, W. Wang and T. Qiao, “An edge detection method in DCT domain”, IEEE ICCE, Las Vegas, Jan. 2017. DCT24 M. Wien, “High Efficiency Video Coding: Coding Tools and Specification”, Springer, 2014. See section 8.1.1 pages 206-210. This section explains how the integer DCTs adopted in HEVC are derived. DCT-P1 Fong, Han and Cham [DCT18] have developed recursive integer cosine transform (RICT). This has orthogonal basis vectors and also has recursive property. Based on this, flow graph for order 32 RICT with flow graphs for orders 4, 8 and 16 embedded is shown in Fig.1. They also have implemented the RICT in HEVC HM13.0 and have demonstrated that the RICT has similar coding efficiency as the core transform in HEVC (see Tables VI thru XI). Using Fig.1, write down the sparse matrix factors (SMFs) for orders 4, 8, 16 and 32 RICTs. DCT-P2 See DCT-P1. Using Table I and the SMFs for orders 4, 8, 16 and 32 RICTs write down the corresponding transform matrices. Verify the transform matrix for order 32 with that shown in eq. 16 of [DCT18]. DCT-P3 See DCT-P1. Forward and inverse architecture of order 4 RICT is shown in Fig.4. This structure is symmetric (Hardware of order 4 inverse RICT can be shared by the order 4 forward RICT.) Develop similar symmetric hardware structures for orders 8, 16 and 32. DCT-P4 The authors in DCT18] state that higher order RICTs such as orders 64 and 128 can be derived using their fast algorithms. Derive these. DCT-P5 See DCT-P4. Draw flow graph for order 64 RICT with the flow graphs for orders 4, 8, 16 and 32 embedded similar to that shown in Fig.1. DCT-P6 Repeat DCT-P5 for order 128 RICT. DCT-P7 See DCT-P6. Extend the suggested values for bN,i for orders 64 and 128 RICTs (See Table I in [DCT18]). DCT-P8 See DCT-P3. Develop similar symmetric hardware structures for orders 64 and 128 RICT. DCT-P9 See Table V in [DCT18]. Extend the comparison to orders 64 and 128. This requires extending Loeffler’s method to orders 64 and 128. DCT-P10 See [DCT11] Jridi and Meher have developed approximate DCT architectures for efficient HEVC compliant video coding. Full-parallel and area constrained architectures for the proposed approximate DCT are described. Show that the equations given by (6a) and (6b) yield the 4-point DCT kernel given by (4).Write down the SMFs based on Fig.1 for the 4-point integer 1-D DCT. DCT-P11 See DCT-P10. Show that the equations given by (7) – (9) yield Eq. (5) for the 8-point DCT kernel. DCT-P12 See DCT-P11. Write down the SMFs for the 8-point integer 1-D DCT based on Fig. 2. Show that these SMFs yield Eq. (5) for the 8-point DCT kernel. DCT-P13 See [DCT11]. Parallel architecture for approximate 8-point 1-D DCT is shown in Fig.2. Develop similar architecture for approximate 16-point 1-D DCT. DCT-P14 Based on DCT-P13 write down the corresponding SMFs and the resulting 16x16 1-D DCT matrix. DCT-P15 See DCT-P13. Extend this to approximate 32-point 1-D DCT. DCT-P16 See DCT-P14 Extend this to 32-point 1-D DCT. DCT-P17 Chen, Han and Cham [DCT21] have developed a set of low complexity integer cosine transforms (LCICTs) and used them in HEVC HM13. These have fully factorizable structures and high circuit reusability. They conclude that order-64 ICTs can significantly improve the coding performance under high QP (low bitrate) which makes them beneficial in low bitrate applications such as video conferencing and video surveillance. Using equations 5-10, write down the LCITC matrices for order 8, 16, 32 and 64. DCT-P18 See DCT-P17 Using Fig.3 (flow graph) write down the SMFs for LCICTs of order 8, 16, and 32. DCT-P19 See DCT-P17. Extend the flow graph for LCICT of order 64. DCT-P20 See DCT-P17 Write down the SMFs for LCICT of order 64. DCT-P21 See [DCT21]. In Table IV the BD bitrates for LCICT are compared with those for another order 64 ICT (see reference 14 at the end) using various test sequences under AI, RA, and LDB. Confirm these results. DCT-P22 The flow graph for order-32 LCICT (flow graphs for order-8 1nd 16 LCICT are embedded) is shown in Fig. 1 [DCT21]. Draw the flow graph for the inverse LCICTs. DCT-P23 Based on DCT-P22, write down the corresponding SMFs for these inverse LCICTs. DCT-P24 See DCT-P22 and DCT-P23. Evaluate the orthogonality of these LCICTs. Hint: matrix multiply the forward LCICT with the matrix for corresponding inverse LCICT. Are these matrices are truly orthogonal? DCT-P25M. Wien, “High Efficiency Video Coding: Coding Tools and Specification”, Springer, 2014. See section 7.3.3 Determination of the interpolation filter (IF) coefficients pages 194-196 for the theory behind how these IF coefficients are derived using DCT. Derive these IF coefficients independently and confirm that they match with the coefficients listed in Tables 5.4 and 5.5. Plot the corresponding transfer functions (See Figs. 7.16 and 7.17 – pages 201-202).
Yüklə 1,18 Mb. Dostları ilə paylaş:
|