在即時視訊傳輸環境中，如何減少傳輸時所造成的延遲是一個非常重要的課題。因為資料流在壓縮編碼器的緩衝區中停留時間必須很短，一個可以避免緩衝區位元率的溢位並同時維持高品質畫面的位元率控制演算法是必要的，但H.264中的位元率控制對於緩衝區位元率的溢位問題並不是處理的很好，往往會造成畫面跳張的問題。
因此我們提出一個新的H.264 位元率的演算法。我們首先以一個根據壓縮過後的I、P、B畫面平均品質來校正I、P、B加權因子方法來估測畫面的初始目標位元數，以便達到I、P、B之間的畫面品質比較平穩，然後再以比例-積分-微分緩衝區控制進一步調整畫面的目標位元數使緩衝區累積位元數接近我們預設的緩衝區大小的一半。一旦我們得到畫面的目標位元數，我們在移動估測之前用R-D二次式去估測畫面的量化參數。在得到估測的畫面量化參數之後，應用Kalman濾波器去進一步動態校正Macroblock的量化參數來讓壓縮位元非常接近目標的位元同時提高畫面的品質並且有效管理緩衝器裡面的累積位元數來減少延遲時間。在跟JVT-G012比較過後，我們的位元率控制演算法能在低位元率的頻寬下可以達到較佳的影像品質而且可以避免緩衝區的上溢/下溢。我們的架構跟JVT-G012比較起來平均PSNR增益要好0.12 dB而且由於應用Kalman濾波器準確的估測Macroblock量化參數所以可以達到壓縮位元數非常接近目標位元數。

In this thesis, we propose a new algorithm for H.264 rate control. We first estimate the frame target bit based on the proportional-integral-derivative (PID) buffer control to avoid the buffer overflow or underflow, and then propose an adaptive method to adjust the weighted factors of the frame target bit estimation. Once we obtain the frame target bit, we use the quadratic R-D model to estimate the frame quantization parameter (QP) before motion estimation. After estimating the frame QP, we apply Kalman filter to further dynamically adjust macroblock QP to achieve accurate bit rate while maximizing the picture quality and simultaneously effectively handing buffer fullness. In comparison with JM7.6, our scheme can give better video quality, reduce sharp drops of PSNR and decreases buffer overflow/underflow even through low bit rate. Most of all, our scheme also achieves an average PSNR gain of 0.127dB and generates coding bits close to target bits due to the accurate macroblock quantization parameter of Kalman filter estimator.

Reference
[1] Z. Li, F. Pan, K. P. Lim, G. Feng, X. Lin and S. Rahardja, “Adaptive basic unit layer rate control for JVT”, JVT-G012-r1, 7th Meeting, Pattaya II, Thailand, Mar. 2003.
[2] T. H. Chiang and Y. Q. Zhang, “A New Rate Control Scheme Using Quadratic Rate Distortion Model”, IEEE Transactions on Circuit and System for Video Technology, Vol.7, No. 1, pp. 246-250, Feb. 1997
[3] F. Pan, Z. Li, K. Lim, and G. Feng, “A study of MPEG-4 rate control scheme and its improvements,” IEEE Trans. on CAS for VT, vol. 13, 440-446, 2003.
[4] Video Group, MPEG-4 Video Verification Model Version 18.0, in Coding of Moving Pictures and Associated Audio MPEG 2001/N3908, Pisa, Italy, Jan. 2001.
[5] Y. Sun and I. Ahmad,”A Robust and Adaptive Rate Control Algorithm for Object-Based Video Coding”, IEEE Transactions on CAS for VT, Vol. 14, No. 10, Oct. 2004
[6] Hyun Mun Kim,” Adaptive Rate Control Using Nonlinear Regression”, IEEE Transactions on CAS for VT, Vol. 13, No. 5, May 2003
[7] Hongtao Yu, Zhiping Lin,” An Improved Rate Control Algorithm for H.264”, IEEE ISCAS, Kobe Japan ,May 2005.
[8] Ping Li, X.K. Yang, W.S. Lin,” Buffer-Constrained R-D Model–Based Rate Control For H.264/Avc”, IEEE ICASSP Philadelphia USA, March 2005.
[9] M. Jiang, X. Yi, and N. Ling,” Improved Frame-Layer Rate Control For H.264 Using Mad Ratio”, IEEE ISCAS China, May 2004.
[10] S. Kim and Yo-Sung Ho, “Rate Control Algorithm for H.264/AVC Video Coding Standard Based on Rate-Quantization Mode”, IEEE ICME Taipei Taiwan, June 2004.
[11] State Key Lab on Microwave & Digital Communications, “A Novel Rate Control For H.264”, IEEE ISCAS China, May 2004.
[12] T.Wiegand, G. J. Sullivan, G. Bjøntegaard, and A. Luthra, “Overview of the H.264/AVC Video Coding Standard,” IEEE Transactions on Circuits And Systems For Video Technology, Vol. 13, No. 7, July 2003
[13] T. Wiegand, H. Schwarz, A. Joch, F. Kossentini, and G. J. Sullivan, “Rate-Constrained Coder Control and Comparison of Video Coding Standards,” IEEE Transactions on CAS for Video Technology, Vol. 13, No. 7, July 2003
[14] E. Parzen, “On Estimation of a Probability Density Function and Mode,” Ann. Math. Statist., vol. 33, pp. 1065–1076, 1962
[15] T. Cacoullos, “Estimation of a Multivariate Density,” Annals Inst. Statist. Math. (Tokyo), vol. 19, no. 2, pp. 179–189, 1966.
[16] S. Haykin, ”Kalman Filtering and Neural Networks”, A Wilet-Interscience Publication, John Wiley & Sons, Inc.