移動估計(motion estimation)在移動補償轉換編碼(MCTC)之視訊壓縮系統中佔有很重要的地位。在此系統中，相鄰影像之間的時間餘贅以移動估計及移動補償技術移除，而轉換編碼則用來消除補償誤差的空間餘贅。區塊移動演算法為最主要的移動估計技術，在執行區塊移動演算法時，影像先分割為相同大小，彼此不重複的區塊，每個區塊各自在參考區域中尋找其移動向量。
在本論文中，我們探討兩個將區塊移動演算法應用在視訊壓縮系統中的問題。首先為如何將可變區塊區塊移動演算法應用在超低位元率壓縮法中；其次則為區塊移動演算法的軟、硬體加速問題。

大多數的以MCTC 為基礎的視訊編碼器皆以 DCT 來移除移動補償誤差影像的空間餘贅，在超低位元率的環境下，由於DCT所分配的位元數極低，使得DCT無法有效地去除空間餘贅。因此，我們發展了一新的視訊壓縮系統來增進MCTC壓縮法的效率。在本論文中，我們從 rate-distortion 的觀點，以理論分析、比較了可變區塊區塊移動演算法（VSBM）與DCT的效率，分析結果顯示，在超低位元率的環境下，VSBM 能較DCT有更加的壓縮效率。另外，我們亦分析了VSBM的各種移動向量分割式樣的效率，並依此提出了限制型VSBM分割法。以此限制型VSBM分割法為基礎，我們提出兩種超低位元率的視訊壓縮法，實驗結果顯示我們的方法較 H.263 有效率。

在區塊移動演算法的硬體加速問題上，我們提出了一個可同時執行全域搜尋區塊比對(FSBM)、2-D小波轉換(wavelet transform)、全域搜尋向量量化(FSVQ)的整合型硬體架構。首先我們以矩陣向量形式來說明三種運算的相似性，以此相似性為基礎，提出了整合型 systolic array 架構。此架構包含了 1-D 的 處理單元矩陣、32 個 cyclic shift 暫存器、4個 delay 暫存器及4個最小誤差偵測器。我們的架構可執行區塊大小16´16、搜尋範圍(-8,7) 的 FSBM；兩階層的 2-D Harr 轉換；區塊大小2´2的FSVQ。

在區塊移動演算法的軟體加速問題上，我們提出了一個 complexity-distortion 最佳化的快速區塊比對法。我們將快速區塊比對問題視為 rate-distortion 最佳化的問題，以此為基礎提出了幾項技術以達到 complexity-distortion 的最佳化。我們以dominate-based 移動向量預估法來設定每個區塊的初始移動向量；predictive complexity-distortion benefit list 用來預估每個區塊的補償利益；adaptive diamond searching 及 integral projection condition checking 則用來搜尋移動向量。我們所提的方法不但比傳統的區塊比對法更有效率，同時亦提供了一個可允許使用者任意調整 computational complexity 的有彈性的移動估計法。

Motion estimation plays an important role in motion compensation transformation coding (MCTC) video compression systems. The temporal redundancy among adjacent frames is removed by motion estimation and motion compensation techniques. Transformation coding reduces the spatial redundancy after compensation. Block matching algorithm (BMA) is the most popular method for motion estimation. For performing BMA, the input image is partitioned into non-overlapped fixed size blocks and each block determines the motion vector in a referenced search area.
In this dissertation, two issues of adopting the blocking matching techniques on motion compensation transformation coding systems are addressed. The first one is the adaptation of BMA to fit into the very low bit rate communication environment. The other is the real-time processing problem of using BMA.

Most MCTC based video coding algorithms employ DCT to reduce spatial redundancy after performing motion compensation. At very low bit rate, the performance of the DCT may become poor because the reserved bit-rate for DCT is limited so it is unlikely to compress the spatial information effectively in that case. Thus, we developed a new approach that could improve the performance of the traditional MCTC methods in the very low bit rate communication environment. At first, a theoretical rate-distortion analysis of the coding efficiency for variable size block matching (VSBM) and DCT is presented. The analysis result shows that employing VSBM can achieve better performance than applying DCT at very low bit rate. We also analyze the rate-distortion performance of different combination of the child motion vectors for VSBM. Based on the analysis result, two algorithms are developed. The resulting performance has an overall improvement compared to H.263.

The second issue of the dissertation is the real-time processing problem of the BMAs. The full-search block matching (FSBM) is the optimal BMA but with large amount of computation. To meet the real-time processing requirement, custom hardware devices or fast BMA should be required. We developed a unified systolic array architecture for operating FSBM, discrete wavelet transform, and full-search vector quantization. The architecture contains one-dimensional processing element array, 32 cyclic shift registers, 4 delay registers and 4 minimum distortion detectors. The FSBM with block size 16´16 and the search range (-8,7), the 2-D 2 level Harr transform with block size 8´8; and the full-search vector quantization with input vector size 2´2, can be efficiently executed in the proposed architecture.

Besides, we also study the software solution of fast BMAs. A complexity-distortion-based optimal fast block matching algorithm is presented. In order to approach the complexity-distortion optimization solution as close as possible, some strategies are developed. A dominate-based motion vector prediction technique is developed to set up the initial motion vector for each block. A predictive complexity-distortion benefit list is established to predict the compensated benefit for each block. An adaptive diamond searching and integral projection condition checking is employed to check the candidate motion vector. The proposed algorithm not only improves the efficiency of the traditional BMAs, but also provides a flexible motion estimation tool that allows user to terminate motion estimation at any computational complexity.

[1] Information Technology-Coding of Moving Pictures and Associated Audio for Digital Storage Media at up to about 1.5 Mbit/s-Part 2: Video, ISO/IEC 11172-2 (MPEG-1 Video),1993.
[2] Information Technology-Generic Coding of Moving Pictures and Associated Audio Information: Video, ISO/IEC 13818-2 (MPEG-2 Video)ITU-T H. 262, 1996.
[3] ITU-T Recommendation H.263, “Video codec for low bitrate communication,” May 1996.
[4] MPEG-4 Video Group, Generic coding of audio-visual objects Part 2 – Visual, ISO/IEC JTC1/SC29/WG11 N1902, FDIS of ISO/IEC 14496-2, Atlantic City, Nov. 1998.
[5] Koga, et al., "Motion compensated interframe coding for video conferencing," Proc. Nat'l. Tele. Conf. Trans. Communications, vol. COM-29, no, 12, pp. 1799-1808, Dec. 1981.
[6] Ya-Qin Zhang, and Sohail Zafar, "Predictive block-matching motion estimation for TV coding--part II: inter-frame prediction," IEEE Trans. on Broadcasting, vol. 37, no. 3, pp. 102-105, Sept. 1991.
[7] M. Bierling, “Displacement estimation by hierarchical blockmatching,” in Proc. SPIE Conf. Visual Communications and Image Processing, SPIE 1988, vol. 1991, pp.942-951.
[8] A. Puri, H. M. Hang, and D. L. Schilling, "Interframe coding with variable block-size motion compensation," in GLOBECOM'87, pp. 65-69, Nov. 1987.
[9] M. H. Chan, Y. B. Yu, and A. G. Constantinides, "Variable size block matching motion compensation with application to video coding," Proc. IEE, Pt. I, vol. 137, pp. 205-212, Aug. 1990.
[10] G. J. Sullivan, and R. L. Baker, "Efficient quadtree coding of images and video," IEEE Trans. Image Processing, vol. 3, pp.327-331, May 1994.
[11] Hao Bi, and Wai-Yip Chan, "Rate-Distortion Optimization of Hierarchical Displacement Fields," IEEE Trans. On Circuits and System for Video Technology, vol. 8, no. 1, pp. 18-25 Feb. 1998.
[12] J. W. Kim, and S. U. Lee, "On the hierarchical variable block size motion estimation for motion sequence coding," Optical Engineering, vol. 33, pp. 2553-2561, Aug. 1994.
[13] H. Musmann, M. Hotter, and J. Ostermann, "Object-oriented analysis-synthesis coding of moving images," Signal Processing:Image Communications, vol. 1, pp. 117-138, Oct. 1989.
[14] N. Diehl, "Object-oriented motion estimation and segmentation in image sequences," Signal Process: Image Communications, vol. 3, pp. 23-56, Feb. 1991.
[15] Michael C. Chen, and Alan N. Willson, Jr. “Rate-Distortion optimal Motion Estimation Algorithms for Motion-Compensated Transform Video Coding,” IEEE Trans. On Circuits and System for Video Technology, vol. 8, no. 2, pp. 147-158 Apr. 1998.
[16] Dimitrios Tzovaras, and Michael G. Strintzis, “Motion and Disparity Field Estimation Using Rate-Distortion Optimization,” IEEE Trans. On Circuits and System for Video Technology, vol. 8, no. 2, pp. 171-180 Apr. 1998.
[17] Liang-Jin Lin, and Antonio Ortega, “Bit-Rate Control Using Piecewise Approximated Rate-Distortion Characteristics,” IEEE Trans. On Circuits and System for Video Technology, vol. 8, no. 4, pp. 446-459 Aug. 1998.
[18] Guido M. Schuster, and Aggelos K. Katsaggelos, “A Video Compression Scheme with Optimal Bit Allocation Among Segmentation, Motion, and Residual Error,” IEEE Trans. On Image Processing, vol. 6, no. 11, pp. 1487-1502, Nov. 1997.
[19] Jiajun Zhang, M. Omair Ahmad, and M. N. S. Swamy, “Quadtree Structured Region-Wise Motion Compensation for Video Compression,” IEEE Trans. On Circuits and System for Video Technology, vol. 9, no. 5 AUG. 1999, pp. 808-822.
[20] Steve Calzone, et. cl., “Video Compression by Mean-Corrected Motion Compensation of Partial Quadtrees,” IEEE Trans. On Circuits and System for Video Technology, vol. 7, no. 1, FEB. 1997, pp. 86-96.
[21] Hsueh-Ming Hang, and Jiann-Jone Chen, "Source Model for Transform Video Coder and Its Application-Part I: Fundamental Theory," IEEE Trans. On Circuits and System for Video Technology, vol. 7, no. 2, pp. 287-298, Apr. 1997.
[22] A. N. Netravali, and B. G. Haskell, Digital Pictures: Representation and Compression. New York, NY: Plenum, 1988.
[23] Thomas Komarek, and Peter Pirsch, "Array architecture for block matching algorithms," IEEE Trans. on Circuits and Systems, no. 10, pp. 1301-1308, Oct. 1989.
[24] Kun-Min Yang, Ming-Ting Sun, and Lancelot Wu, "A family of VLSI designs for the motion compensation block-matching algorithm," IEEE Trans. on Circuits and Systems, vol. 36, no. 10, pp. 1317-1325, Oct. 1989.
[25] Chen-Mie Wu, and Ding-Kuen Yeh, "A VLSI motion estimation for video image compression," IEEE Trans. on Consumer Electronics, vol. 39, no. 4, Nov. 1993.
[26] Seung Hyun Nam, Jong Seob Baek, and Moon Key Lee, "Flexible VLSI architecture of full search motion estimation for video applications," IEEE Trans. on Consumer Electronics, vol. 40, no. 2, pp. 176-183, May 1994.
[27] Viet L. Do, and Kenneth Y. Yun, “A low-power VLSI architecture for full-search block-matching motion estimation,” IEEE. Trans. on Circuits Syst. Video Technol., Vol. 8, no. 4, pp. 393-398, Aug. 1998.
[28] Sheu-Chih Cheng, and Hsueh-Ming Hang, "A Comparison of Block-Matching Algorithms Mapped to Systolic-Array Implementation," IEEE. Trans. on Circuits Syst. Video Technol., vol. 7, no. 5, pp. 741-753, Oct. 1997.
[29] Yuan-Hau Yeh, and Chen-Yi Lee, "Cost-Effective VLSI Architectures and Buffer Size Optimization for Full-search Block Matching Algorithms," IEEE Trans. on VLSI Systems, vol. 7, no. 3, pp. 345-358, Sep. 1999.
[30] Yeong-Kang Lai, and Liang-Gee Chen, "A Data-Interlacing Architecture with Two-Dimensional Data-Reuse for Full-Search Block-Matching Algorithm," IEEE. Trans. on Circuits Syst. Video Technol., vol. 8, no. 2, pp.124-127, Apr. 1998.
[31] Alexander Bugeja, and Woodward Yang, "A Reconfigurable VLSI Coprocessing System for the Block Matching Algorithm," IEEE Trans. on VLSI Systems, vol. 5, no. 3, pp. 329-337, Sep. 1997.
[32] Yunju Baek, Hwang-Seok Oh, and Heung-Kyu Lee, "An Efficient Block-Matching Criterion for Motion Estimation and Its VLSI Implementation," IEEE Trans. on Consumer Electronics, vol. 42, no. 4, pp. 885-892, Nov. 1996.
[33] Yeu-Shen Jehng, Liang-Gee Chen, and Tzi-Dar Chiueh, "An Efficient and Simple VLSI Tree Architecture for Motion Estimation Algorithms," IEEE Trans. on Signal Processing, vol. 41, no. 2, pp. 889-900, Feb. 1993.
[34] Aleksander Grzeszczak, Mrinal K. Mandal, Sethuranman Panchanathan, and Tet Yeap, “ VLSI implementation of discrete wavelet tansform,” IEEE Trans. on VLSI Syst., vol. 4, no. 4, pp. 421-433, DEC. 1996.
[35] Keshab K. Parhi, and Takao Nishitani, “VLSI architectures for discrete wavelet transforms,” IEEE Trans. on VLSI Syst., vol. 1, no. 2, pp. 191-202, June 1993.
[36] Gauthier Lafruit, Francky Catthoor, Jan P. H. Cornelis, and Hugo J. De Man, “An efficient VLSI architecture for 2-D wavelet image coding with novel image scan,” IEEE Trans. on VLSI Syst., vol. 7, no. 1, pp. 56-68, March 1999.
[37] Chaitali Chakrabarti, and Clint Mumford, "Efficient Realizations of Encoders and Decodes Based on the 2-D Discrete Wavelet Transform," IEEE Trans. on VLSI Systems, vol. 7, no. 3, pp. 289-298, Sep. 1999.
[38] Mohan Vishwanath, Robert Michael Owens, and Mary Jane Irwin, "VLSI Architectures for the Discrete Wavelet Tranform," IEEE Trans. on Circuits and Systems-II: Analogy and Digital Signal Processing, vol. 41, no. 5, pp. 305-316, May 1995.
[39] Henry Y. H. Chuang, and Ling Chen, "VLSI Architecture for Fast 2D Discrete Orthonormal Wavelet Transform," Journal of VLSI Signal Processing, 10, pp. 225-236, 1995.
[40] Chu Yu, and Sao-Jie Chen, "Design of an Efficient VLSI Architecture for 2-D Discrete Wavelet Transforms," IEEE Trans. on Consumer Electronics, vol. 45, no. 1, pp. 135-140, Feb. 1999.
[41] Tracy C. Denk, and Keshab K. Parhi, "VLSI Architectures for Lattice Structure Based Orthonormal Discrete Wavelet Tranforms," IEEE Trans. on Circuits and Systems-II: Analogy and Digital Signal Processing, vol. 44, no. 2, pp. 129-132, Feb. 1997.
[42] Chaitali Chakrabarti, Mohan Vishwanath, and Robert M. Owens, "Architectures for Wavelet Transforms: A Survey," Journal of VLSI Signal Processing, 14, pp. 171-192, 1996.
[43] Chin-Liang Wang, and Ker-Min Chen, “A New VLSI Architecture for Full-Search Vector Quantization,” IEEE. Trans. on Circuits Syst. Video Technol., vol. 6, no. 4, pp. 389-398, Aug. 1996.
[44] Wai-Chi Fang, Chi-Yung Chang, Bing J. Sheu, Oscal T.-C. Chen, and John C. Curlander, “VLSI Systolic Binary Tree-Searched Vector Quantizer for Image Compression,” IEEE Trans. on VLSI Syst., vol. 2, no. 1, pp. 33-44, March 1994.
[45] S. G. Mallat, “A theory for multiresolution signal decomposition: The wavelet representation,” IEEE Trans. Pattern Anal. Machine Intell., vol. 11, pp. 674-693, July 1989.
[46] H. G. Musmann et al., "Advances in picture coding," Proc. IEEE 73, pp. 523-548, 1985.
[47] R. M. Gray, "Vector quantization," IEEE ASSP Mag., vol. 1, no. 2, pp. 4-29, Apr. 1984.
[48] Kung, H. T. "Why systolic architectures?," Computer, pp. 37-45, Jan. 1982.
[49] Michael C. Chen, and Alan N. Willson, Jr., “Rate-distortion optimal motion estimation algorithms for motion-compensated transform video coding,” IEEE Trans. Circuits Syst. Video Tech., vol. 8, no. 2, pp. 147-158, Apr. (1998).
[50] K. Cheung, and L. M. Po, “Normalized partial distortion search algorithm for block motion estimation,” IEEE Trans. Circuits Syst. Video Tech., vol. 10, no. 3, pp. 417-422, Apr. (2000).
[51] T. Koga, K. Iinuma, A. Iijima, and T. Ishiguro, “Motion-compensated interframe coding for video conderencing,” in Proc. NTC81, New Orleans, LA, pp. C9.6.1-9.6.5, 1981.
[52] Faouzi Kossentini, Yuen-Wen Lee, Mark J. T. Smith, and Rabab K. Ward, “Predictive RD optimized motion estimation for very low bit-rate video coding,” IEEE Journal on Selected Areas in Communications, vol. 15, no. 9, pp. 1752-1763, Dec. (1997).
[53] Liu and, A. Zaccarin, “New fast algorithms for the estimation of block motion vectors,” IEEE Trans. Circuits Syst. Video Tech., vol. 3, pp. 148-157, Apr. (1993).
[54] Yih-Chuan Liu; and Shen-Chuan Tai, “Fast full-search block-matching algorithm for motion-compensated video compression,” ICPR’96, vol. 3, pp.914-918.
[55] L. M. Po, and W. C. Ma, “A novel four-step search algorithm for fast block motion estimation,” IEEE Trans. Circuits Syst. Video Tech., Vol. 6, no. 3, pp. 313-317, (1996).
[56] R. Li, B. Zeng, and M. L. Liou, "A new three-step search algorithm for block motion estimation," IEEE Trans. Circuits Syst. Video Tech., vol. 4, pp.438-442, Aug. 1994.
[57] Vassilios Christopoulos, and Jan Cornelis, "A Center-Biased Adaptive Search Algorithm for Block Motion Estimation," IEEE Trans. Circuits Syst. Video Tech., vol. 10, no. 3, pp. 423-426, Apr. 2000.
[58] Michael Gallant, Guy Côté, and Faouzi Kossentini, "An Efficient Computation-Constrained Block-Based Motion Estimation Algorithm for Low Bit Rate Video Coding," IEEE Trans. Image Processing, vol. 8, no. 12, pp. 1862-1823, Dec. 1999.
[59] Oscal T.-C. Chen, "Motion Estimation Using a One-Dimensional Gradient Descent Search," IEEE Trans. Circuits Syst. Video Tech., vol. 10, no. 4, June 2000.
[60] Chun-Ho Cheung, and Lai-Man Po, "Generalized Partial Distortion Search Algorithm for Fast Block Motion Estimation," ICASSP2001.
[61] Bing Zeng, Renxiang Li, and Ming L. Liou, "Optimization of Fast Block Motion Estimation Algorithms," IEEE Trans. Circuits Syst. Video Tech., vol. 7, no. 6, pp. 833-844, Dec. 1997.
[62] Yankang Wang, and Hideo Kuroda, "Hilbert Scanning Search Algorithm for Motion Estimation," IEEE Trans. Circuits Syst. Video Tech., vol. 9, no. 5, pp. 683-691, Aug. 1999.
[63] Jie-Bin Xu, Lai-Man Po, and Chok-Kwan Cheung, "A new prediction model search algorithm for fast block motion estimation," ICIP, vol. 3, pp. 610-613, 1997.
[64] K. W. Cheng, and S. C. Chan, "Fast block matching algorithms for motion estimation," ICASSP-96, vol. 4, pp.2311-2314, 1996.
[65] H. T. Tang, J. Feng, and K. T. Lo, "New fast search algorithm for block matching motion estimation," ICIP99, vol. 3, pp. 681-685, 1999.
[66] C. F. Chang, and J. S. Wang, "A stable buffer control strategy for MPEG coding," IEEE Trans. Circuits Syst. Video Tech., vol. 7, pp.920-924, Dec. 1997.
[67] Yoonho Kim, and Hansoo Kim, "A new buffer control strategy for image data compression," IEEE Trans. Consumer Electronics, vol. 40, no. 4, pp. 932-937, Nov. 1994.
[68] Hwang-Seok Oh, and Heung-Kyu Lee, "Adaptive rate control scheme for very low bit rate video coding," IEEE Trans. Consumer Electronics, vol. 42, no. 4, pp. 974-980, Nov. 1996.
[69] Pedro A. A. A., and Mohammed Ghanbari, " A buffer control algorithm for CBR video transcoding," ICIP98, vol. 2, pp. 370-374, 1998.
[70] K. T. Ng, S. C. Chan, and T. S. Ng, "Buffer control algorithm for low bit-rate video compression," ICIP96, vol. 1, pp. 685-688, 1996.
[71] Injong Rhee, Graham R. Martin, S. Muthukrishnan, and Roger A. Packwood, "Quadtree-Structured Variable-Size Block-Matching Motion Estimation with Minimal Error," IEEE Trans. Circuits Syst. Video Tech., vol. 10, no. 1, pp. 42-50, Feb. 2000.
[72] Yan yang, and Sheila S. Hemami, "Generalized Rate-Distortion Optimization for Motion-Compensated Video Coders," IEEE Trans. Circuits Syst. Video Tech., vol. 10, no. 6, pp. 942-955, Sep. 2000.
[73] J. Lee, "Optimal quadtree for variable block size motion estimation," ICIP, vol. 3, pp. 480-483, Oct. 1995.