臺灣博碩士論文加值系統

由於多媒體應用和網路的快速發展，多媒體通訊變得越來越受歡迎。為了有利於網路傳輸，H.264/AVC 編碼器通常把整張畫面的位元串流放進一個網路封包裡，因此視訊內容在容易出錯的傳輸通道中受到遺失或破壞，在解碼器中所解碼出來的視訊內容將會嚴重地降低視覺品質，為了避免這個問題發生，錯誤隱藏的方法被廣泛的用來回復遺失的畫面並防止錯誤傳播。
在本篇論文中，我們提出一個適應性搜尋範圍去精煉移動向量的演算法，用來改善常用的複製移動向量(motion copy)錯誤隱藏方法，因此提高了錯誤隱藏效果。
我們的方法利用受損宏區塊周邊的移動向量決定ㄧ個適合的搜尋範圍，並利用邊緣的資訊決定是否需要切成不同區塊大小去修復。在高變動的宏區塊時所使用的搜尋範圍較大，反之，在低變動的宏區塊時所使用的搜尋範圍較小，如此就能降低運算複雜度的同時兼顧影像品質。
實驗結果顯示我們提出的方法比其他錯誤補隱藏演算法佳，在客觀品質的峰值訊號雜訊比有明顯的上升，且主觀品質上畫面有明顯的改善。

Owing to the rapid development of multimedia applications and networks, the multimedia communication has become more and more popular. To benefit the network transmission, H.264/AVC encoder usually codes the whole frame into single packet. Therefore, if video contents suffer from the packet loss in error-prone transmission channel, the visual quality will seriously degrade in decoder. To deal with this problem, error concealment method is widely utilized to recover the lost frame and prevent the error propagation.
In this thesis, we propose a motion vector refinement technique with an adaptive search range to further refine the well-known motion copy error concealment method and thus improve the concealed results. By using the surrounding motion vectors of the corrupted marcoblock, we can decide an adaptive search range. In addition, different block sizes are classified by edge information in order to enhance accuracy of motion vectors. High-motion-variation macroblock has larger search range. On the contrary, low-motion-variation macroblock has smaller search range. Therefore, our algorithm can reduce a lot of computation complexity and increase visual quality in the same time.
Simulation results show that our proposed algorithm can achieve significant gains in peak-to-peak signal to noise ratio and improve subjective quality effectively when compared to conventional schemes.

ABSTRACT
ACKNOWLEDGMENT
CONTENTS
LIST OF TABLES
LIST OF FIGURES

Chapter 1
Introduction 1
1.1 H.264/AVC Video Compression Standard 1
1.2 Motivation 3
1.3 Thesis Organization 4
Chapter 2
Background 6
2.1 Overview of Error Concealment and Error Resilience 6
2.2 The Slice Structuring in H.264 Encoder 7
2.3 The Error Concealment in H.264 Decoder 9
2.4 The GOP Structure 10
2.5 Error Propagation 11

Chapter 3
Related Works 12
3.1 Frame Copy 12
3.2 Motion Copy 13
3.3 Boundary Matching Algorithm (BMA) 14
3.4 Extend Boundary Matching Algorithm (EBMA) 16
3.5 Block Size Selection (BSS) 18
3.6 Motion Vector Refinement Scheme 19
3.7 Distance functions 22
3.7.1 Sum of squared differences (SSD) 22
3.7.2 Sum of absolute differences (SAD) 23

Chapter 4
Proposed Algorithm and Architecture Implementation 24
4.1 Preliminary Analysis 24
4.2 Proposed Algorithm 26
4.2.1 Block Size Selection 28
4.2.2 Adaptive Search Range Decision 32
4.3 Hardware Implementation 36
4.3.1 EC Processing Element 37
4.3.2 Frame Register 39
4.3.3 64 Processing Elements 41
4.3.4 SAD Tree 42

Chapter 5
Experimental Results 43
5.1 Experimental Platform 43
5.1.1 Test Sequences 44
5.1.2 Error Pattern 44
5.2 Comparison of PSNR with Other Methods 45
5.3 Comparison of Subjective Quality 47
5.4 Implementation 68
5.4.1 Timing Report 68
5.4.2 Area Report 70
5.4.3 Total Report with tsmc 0.18um Library 71
Chapter 6
Conclusions 72

References 73

[1] Wiegand, T.; Sullivan, G.J., “The H.264/AVC Video Coding Standard, IEEE Signal Processing Magazine, vol:24, pp.148-153, March 2007
[2] Y. C. Chen, W. G. Wu, “Design and Analysis of H.264 Intra-Prediction and Mode Decision Architecture Using Rate-Distortion Optimization, July 2006
[3] Detlev Marpe and Thomas Wiegand, Heinrich Hertz Institute (HHI),Gary J. Sullivan, Microsoft Corporation, “The H.264/MPEG4 Advanced Video Coding Standard and its Applications, IEEE Communications Magazine, vol:44, PP. 134-143, August 2006
[4] ITU-T Recommendation H.264, Advanced video coding for generic audiovisual services, May 2003.
[5] Joint Video Team, H.264/AVC Reference Software JM16.2:
http://iphome.hhi.de/suehring/tml/download/
[6] W.-M. Lam, A. R. Reibman and B. Liu, “Recovery of lost or erroneously received motion vectors, IEEE International Conference on Acoustics, Speech, and Signal Processing, Vol. 5, pp. 417-420, Apr. 1993.
[7] Thaipanich, T. ; Ping-Hao Wu ; Kuo, C.-C.J., “Low-Complexity Mobile Video Error Concealment Using OBMA, Consumer Electronics, 2008. ICCE 2008. Digest of Technical Papers. International Conference on, Jan. 2008.
[8] M. Yang and N. G. Bourbakis, “An Efficient Packet Loss Recovery Methodology for Video Streaming Over IP Networks, IEEE Transactions on Broadcasting, vol. 55, pp. 190-201, Jun. 2009.
[9] Wenjie Yang ; Xingang Liu ; Chao Sun, “Improved Motion Vector Recovery in Temporal Error Concealment for H.264/AVC Decoder, Computer and Information Technology (CIT), 2012 IEEE 12th International Conference on, Oct. 2012.
[10] Yan Sun ; Songlin Sun ; Xiaojun Jing ; LiXin Zhao, “A Dynamic Temporal Error Concealment Algorithm for H.264, Multimedia Technology (ICMT), 2010 International Conference on, Oct. 2010.
[11] Lei Wang ; Jun Wang ; Goto, S. ; Ikenaga, T., “Adaptive Motion Vector Retrieval Schemes for H.264 Error Concealment, Image and Signal Processing, 2008. CISP '08. Congress on, May 2008.
[12] Guan-Lin Wu ; Ching-Yi Chen ; Shao-Yi Chien, “Algorithm and Architecture Design of Image Inpainting Engine for Video Error Concealment Applications, Circuits and Systems for Video Technology, IEEE Transactions on, vol. 21, pp. 792-803, June 2011.