臺灣博碩士論文加值系統

MPEG-4中顆粒可尺寸調整之研究

許錦森* 賴源泰**

國立成功大學電機工程研究所

摘 要

多媒體中包含聲音與影像的資料。多媒體在商業化產品與娛樂應用方面都扮演一個很重要的角色。在影像處理與傳輸方面，MPEG-4是目前最流行與熱門的技巧。顆粒可尺寸調整是MPEG-4中的一項特性，它可以在給定的位元率範圍區間內，將影像編碼成兩層或多層的位元資料流，以獲得較佳解析度的影像品質。然而，若我們將所有的巨集區塊進行顆粒尺寸調整的動作，則完成整個畫面影像移動估計的時間將會非常冗長。

在本論文中，我們提出一個有效率的顆粒尺寸調整的架構已取代傳統的顆粒可尺寸調整的架構。我們增加一個名為巨集區塊選擇的功能方塊用以評估一個巨集區塊在相鄰的兩張影像中是否有較大的差異。若巨集區塊差異小時，我們略過該巨集區塊並且不執行移動估計，否則，該巨集區塊必須執行移動估計。藉由巨集區塊選擇功能，我們可以減少一個畫面中所需執行移動估計的巨集區塊數目。在移動估計計算中，我們也提出一個新的區塊對應演算法，名為改良型四步搜尋法。它由四步搜尋法改良而來，其效能接近原本的四步搜尋法但改良型四步搜尋法有更少的平均搜尋點數。最後，在我們的實驗數據中將可以證明所提出的方法是相當有效率的。

* 作者 ** 指導教授

Study on the Fine Granularity Scalability
in MPEG-4

Chin-Sen Hsu* Yen-Tai Lai**

Department of Electrical Engineering
Nation Cheng Kung University
Tainan, Taiwan.

ABSTRACT

Multimedia includes audio and video data. It plays an important role in the commercial products and entertainment applications. MPEG-4 is the one of the most popular techniques in video processing and transferring. Fine granularity scalability (FGS) is one of the properties that can get better resolution in a given range of bit rates by providing two or more layers of bitstreams in MPEG-4. However, if we execute FGS for all macro-blocks, the total motion estimation time will raise largely.

In this thesis, we present an efficient FGS architecture instead of the traditional FGS architecture. We add a new functional block called macro-block selection (MB selection) to evaluate a MB whether this MB has large difference between two succeeding pictures or not. If the difference in this MB is very small, we ignore this MB and do not execute motion estimation with this MB. Otherwise, it is necessary to execute motion estimation for this MB. By using MB selection function, it can reduce the number of macro-blocks that need to execute motion estimation in a frame. In the motion estimation, we propose a new block matching algorithm called the modified four step search algorithm (M4SS algorithm) which is an improved version of the four step search algorithm (4SS algorithm). The M4SS has similar performance with 4SS but it has fewer searching points in average. The experimental results confirm the efficiency of the proposed method.

The author* The advisor**

TABLE OF CONTENTS
Page
TABLE OF CONTENTS………………………………………………………... i
LIST OF TABLE…………………………………………………………………..iii
LIST OF FIGURES………………………………………………………………iv
ABSTRACT…………………………………………………………………………..v

CHAPTER 1. INTRODUCTION……………………………………………1
1.1 Background…………………………………………………..1
1.2 Outline of This Thesis………………………………………...2

CHAPTER 2. Video Processing ……………………………………………...3
2.1 Video Picture Format………………………………………………...3
2.2 Discrete Cosine Transform…………………………………………..6
2.3 Quantization………………………………………………………….7
2.4 Motion Compensation……………………………………………….8

CHAPTER 3. Fine Granularity Scalability……………………………10
3.1 Introduction………………………………………………………...10
3.2 Temporal Scalability………………………………………………..11
3.3 Signal-to-Noise Ratio Scalability in MPEG-2……………………11
3.4 Spatial Scalability………………………………………………….14
3.5 Fine Granularity Scalability Architecture………………………….15
3.6 The Modified Fine Granularity Scalability………………………...16
3.6.1 Macro Block Selection………………………………………18
3.6.2 Threshold Discussion ……………………………………….19
3.7 Experimental Results……………………………………………….21

CHAPTER4. Motion Estimation…………………………………………...26
4.1 Introduction…………………………………………………………26
4.2 Block Matching Algorithms………………………………………26
4.2.1 Three Step Search (TSS) algorithm…………………………..27
4.2.2 Novel Three Step Search (NTSS) algorithm…………………28
4.2.3 Four Step Search (4SS) algorithm……………………………30
4.2.4 Diamond Search (DS) algorithm……………………………31
4.2.5 Modified Four Step Search (M4SS) algorithm…………….33
4.3 Experimental Results……………………………………………….37

CHAPTER 5. Hardware Design of Motion Estimation…………46
5.1 Hardware Architecture……………………………………………..46
5.2 Macro Block Selection Unit………………………………………..47
5.3 Absolute Difference Unit (AD Unit)……………………………….48
5.4 Experimental Results……………………………………………….49

CHAPTER 6. Conclusions…………………………………………………….52

Reference

[1]Thomas Sikora,“The MPEG-4 Video Standard Verification Model,”IEEE Trans. Circuit Syst. Video Technol., vol. 7, pp. 19-31, Feb. 1997.

[2] H. Kataya, N. Ito, and H. Kusao, “Temporal-scalable coding based on iamge content,” IEEE Trans. Circuits Syst. Video Technol., vol. 7, pp. 52-59, Feb. 1997.

[3] G. J. Conklin and S. S. Hemami, “A comparison of temporal scalability techniques,” IEEE Trans. Circuits Syst. Video Technol., vol. 9, pp. 909-919, Sept. 1999.

[4] D. Wilson and M. Ghanbari, “Exploiting interlayer correlation of SNR scalable video,” IEEE Trans. Circuits Syst. Video Technol., vol. 9, pp. 783-797, Aug. 1999.

[5] J. F. Arnold, M. R. Frater, and Y. Wang, “Efficient drift-free signal-to-noise ratio scalability,” IEEE Trans. Circuits Syst. Video Technol., vol. 10, pp. 70-82, Feb. 2000.


[6] Weiping Li, “Overview of Fine Granularity Scalability in MPEG-4 Video Standard,” IEEE Trans. Circuits Syst. Video Technol., vol. 11, pp. 301-317, Mar. 2001.

[7] M. Domanski, A. Luczak, and S. Mackowiak, “Spatial-temporal scalability for MPEG
video coding,” IEEE Trans. Circuits Syst. Video Technol., vol 10, pp. 1088-1093, Oct. 2000.

[8] M. van der Schaar, “A Hybrid Temporal-SNR Fine-Granular Scalability for Internet Video,” IEEE Trans. Circuits Syst. Video Technol., vol.11, pp. 318-331, Mar. 2001.


[9] H. Radha, Y. Chen, K. Parthasarathy, and R. Cohen, “Scalable internet video using MPEG-4,” Signal Processing: Image Commun., no. 15, pp.95-126, Sept. 1999.

[10] M. Ghanbari, “The cross-search algorithm for motion estimation,” IEEE Trans. Commun., vol. 38, pp.950-953, July 1990.

[11] L. W. Lee, J. F. Wang, J. Y. Lee, and J. D. Shie, “Dynamic search window adjustment and interlaced search for block-matching algorithm,” IEEE Trans. Circuits Syst. Video Technol., vol. 3, pp. 85-87, Feb. 1993.

[12] B. Liu and A. Zaccarin, “New fast algorithm for estimation of block motion vectors,” IEEE Trans. Circuits Syst. Video Technol., vol. 3, pp. 148-157, Apr. 1993.

[13] Renxiang Li, Bing Zeng, and Ming L. Liou, “A New Three-Step Search Algorithm for Block Motion Estimation,” IEEE Trans. Circuits Syst. Video Technol., vol.4, pp. 438-442, Aug. 1994.

[14] Lai-Man Po and Wing Chung Ma, “A Novel Four-Step Search Algorithm for Fast Block Motion Estimation,” IEEE Trans. Circuits Syst. Video Technol., vol. 6, pp313-317, Jun. 1996.

[15] J. Y. Tham, S. Ranganath, M. Ranganath, and A. A. Kassim, “A novel unrestricted center-biased diamond search algorithm for block motion estimation,” IEEE Trans. Circuits Syst. Video Technol.,vol. 8, pp. 369-377, Aug. 1998.

[16] Shan Zhu and K. K. Ma, “A New Diamond Search Algorithm for Fast Block-Matching Motion Estimation,” IEEE Trans. Image Processing, vol.9 , pp. 287-290, Feb. 2000.

[17] Y. L. Chan and W. C. Siu, “An Efficient Search Strategy for Block Motion Estimation Using Image Features,” IEEE Trans. Image Processing, vol. 10, pp. 1223-1237, Aug. 2001.
[18] K. H. Lee, J. H. Choi, B. K. Lee, and D. G. Kim, “Fast two-step half-pixel accuracy
motion vector prediction,” IEEE Electronics Letters, vol.36, pp. 625-627, Mar. 2000.
[19] C. M. Wu and D. K. Yeh,”A VLSI Motion Estimator for Video Image Compression,”
IEEE Trans. Consumer Electronics, vol.39, pp. 837-845, Nov. 1993.