MPEG-4 無失真音訊編碼之適應性參數判定法則

PARAMETER ADAPTATION ALGORITHMS FOR MPEG-4 AUDIO LOSSLESS CODING

陳易佐(Yi-Zus Chen)

指導教授 ： 李清坤

大同大學/電機資訊學院/通訊工程研究所/碩士(2006年)

摘要

無失真音訊編碼是MPEG-4音訊編碼群組裡面一個新增訂的標準。在2005年12月27日，由日本NTT電信公司通訊科學研究所的無失真音訊編碼技術提案正式被認可為MPEG-4的增訂標準，並簡稱為“MPEG-4 ALS”。共同合作研究此標準的還包括有德國柏林科技大學、美國RealNetworks公司及新加坡I2R研究中心。這個新興的標準MPEG-4 ALS代表了一個有效率的無失真編碼，可以使壓縮後的數位音訊資料能夠完整的重建回原始信號，當然在聽覺品質上就不會有任何的失真。
在本篇論文中，我們主要研究的工作是對於MPEG-4 ALS標準中開放設計的兩項參數調整機制，亦即區塊長度轉換﹙block length switching﹚和預估器階數調整﹙prediction order adaptation﹚發展出適應性估測機制。關於區塊長度轉換這個參數，經由初步實驗，我們從8種可變區塊長度模式選擇了其中4種較佳的模式，並設計了一個適應性估測機制能達成選取最佳區塊長度達95.12%。我們也提出了一個預估器階數的適應性估測機制，實驗結果顯示，相較於完整搜尋法，我們提出的方法可以選取到最佳前3名預估器階數的百分比達85.83%。最後我們將上述兩個參數適應性估測機制應用於MPEG-4 ALS，實驗結果顯示，相較於固定預估器階數6與固定區塊長度1000 - 4000我們提出的方法可以達成降低平均壓縮比1.42% - 2.35%。唯上述方法所需計算時間仍然偏高，期待在未來的發展上能夠繼續有所改進。

關鍵字

無失真音訊編碼

並列摘要

Audio Lossless Coding is a new extension of MPEG-4 audio coding family. In December 27 2005, NTT Corp. has announced that an audio lossless coding technology was approved as an MPEG international standard specification dubbed“MPEG-4 ALS”. Partners in this standardization work included the Technical University of Berlin (Germany), RealNetworks Corp. (USA), and I2R (Singapore). The new emerging standard MPEG-4 ALS (Audio Lossless Coding) introduces an efficient lossless coding that enables the perfect reconstruction of the original signal and the compression of digital audio data without any loss in quality.
In this thesis our ongoing work develops adaptive estimated algorithms for two open designed parameters (block length switching and prediction order adaptation) in MPEG-4 ALS. About the block length switching, we get four best block length models from eight variable block length switching models from the first test, and design a adaptive estimated algorithm which has 95.12% for selecting best block length. About the prediction order adaptation, we propose a adaptive estimated algorithm which has 85.83% for selecting forward 3 best prediction order compared to the algorithm of prediction order full search from the experiment results. Finally, we apply the two adaptive estimated algorithms (block length switching with prediction order adaptation) in MPEG-4 ALS. Our proposed method can reduce 1.42% - 2.35% average compression ratio compared to the method of prediction order 6 + fixed block length 1000 – 4000. However, the computation load is still high, further improvement is expected.

並列關鍵字

MPEG-4 ALS

參考文獻

[1] L. Xiao, T. W. Hong, G. C. Woei, and L. Gang, “A Lossless Audio Compression Software for Windows Application,” Proc. Signal Processing IEEE ICSP, 1998, pp. 1162-1165.

[2] T. Liebchen and Y. A. Reznik, “MPEG-4 ALS: an Emerging Standard for Lossless Audio Coding,” Proc. IEEE Data Compression Conference, 2004, pp. 439-448.

[11] R. F. Rice and J. R. Plaunt, “Adaptive variable-length coding for efficient compression of spacecraft television data,” IEEE Transactions on Communication Technology, vol. 19, pp. 889-897, 1971.

[13] T. Liebchen, Y. Reznik, T. Moriya, and D. T. Yang, “MPEG-4 Audio Lossless Coding,” 116th AES Convention, 2004, no. 6047, pp. 1-9.

[14] T. Liebchen, “An Introduction to MPEG-4 Audio Lossless Coding,” Proc. IEEE ICASSP, 2004, pp. 1012-1015.