基于概率更新机制的自适应混合熵编码

doi:1000-436X(2008)05-0043-07

摘要/Abstract

摘要：

在熵编码引入了一种概率更新的机制，记录每个语法元素的实时概率分布。将基于概率更新的Huffman 码和结构码字有效结合，并根据各个语法元素的相关性选择最恰当的上下文模型，充分利用视频信号的多样性进行信息压缩。实验表明，此种基于概率更新的混合熵编码(PUAHC)算法的编码效率和编码复杂度皆介于 UVLC 和CABAC之间。

关键词: 信息处理技术, 视频编码, 基于概率更新的混合熵编码, 概率估计, Huffman码

Abstract:

A probability updating mechanism was proposed for entropy coding.To achieve the best coding efficiency,probability updating based huffman code and structured code were jointly used,and the best context model was selected according as the correlation of adjacent syntax elements.Experiment results demonstrate that both of the coding efficiency and computational complexity of this probability updating-based adaptive hybrid coding (PUAHC) were between UVLC and CABAC.

Key words: signal processing technology, video coding, PUAHC, probability estimation, Huffman code

徐士麟,喻莉,朱光喜. 基于概率更新机制的自适应混合熵编码[J]. 通信学报, 2008, 29(5): 43-49.

Shi-lin XU,Li YU,Guang-xi ZHU. Probability updating-based adaptive hybrid coding[J]. Journal on Communications, 2008, 29(5): 43-49.

图/表 10

图1

图2

图3

表1

表2

表3

表4

表5

表6

图4

参考文献 10

[1]	ISO/IEC 13 818-2,part2:Generic Coding of Moving Pictures and Associated Audio,Video[S].
[2]	ISO/IEC JTC1/SC29/WG11 N1796:MPEG-4 Video Verification Model Version 8.0[S].
[3]	ITU-T Recommendation H.263.Video Coding for Low Bit Rate Communication Transmission of non – Telephone Signals[S].
[4]	ITU-T SG16 Experts Group Draft H.263 + .Video Coding for Low Bit Rate Communication Transmission of Non-Telephone Signals[S].
[5]	Joint Video Team (JVT) of ISO/IEC MPEG and ITU-T VCEG,JVTG050:Draft ITU-T Recommendation and Final Draft International Standard of Joint Video Specification (ITU-T Rec.H.264/ISO/IEC 14 496-10 AVC)[S].
[6]	WIEGAND T , SULLIVAN G J , BJONTEGAARD G.Overview of the H . 264/AVC video coding standard[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2003,13(7): 560-576.
[7]	ITOH Y,CHEUNG.N . Universal variable length code for DCT coding[A]. International Conference on Image Processings[C]. Vancouver, 2000. 940-943.
[8]	HARTUNG F , GIROD B . Improved encoding of DCT coefficients for low bit-rate video coding using multiple VLC tables[A]. International Conference on Image Processings[C]. Kobe, 1999. 51-55.
[9]	WANG Q , ZHAO D , MA S . Context-based 2D-VLC for video coding[A]. IEEE International Conference on Multimedia and Expo[C]. Taipei, 2004. 89-92.
[10]	MARPE D , SCHWARZ H , WIEGAND T.Context-based adaptive binary arithmetic coding in the H . 264/AVC video compression standard[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2003,13(7): 620-636.

869	209	263	38	16	17	4	4	3	22	1	1	1	1	1	1
151	253	328	64	203	93	50	72	286	30	54	6	7	3	1
40	174	76	157	73	60	75	314	18	90	14	7	21	2
18	49	45	72	70	87	173	24	76	19	11	13	1
8	35	23	36	57	95	38	75	21	21	17	2
9	8	33	42	64	26	67	27	18	22	3
2	7	25	47	22	64	21	11	28	1
5	10	14	13	24	31	16	22	5
4	4	6	17	15	20	13	3
2	2	11	6	9	22	5
1	4	5	3	9	1
1	4	3	2	1
1	1	4	2
2	1	2
1	1

leftrun	每一个leftrun对应的run的取值范围	语法元素run对应的概率表
≥6	0～14	258，265，269，230，239，423，209， 98，80，29，13，7，2，1，0
5	0～5	229，157，165，153，137，135
4	0～4	301，281，222，220，265
3	0～3	495，435，303，520
2	0～2	1 011，730，1 048
1	0,1	1 627，2 180

level	模式1	模式2
1	1	10
?1	01	11
2	001	010
?2	0001	011
3	00001	0010
?3	000001	0011
4	0000001	00010
?4	00000001	00011
…	…	…

R^*[1][0]	R^*[1][1]	R^*[1][2]	R^*[1][3]	R^*[1][4]	R^*[1][5]
0.234 631	0.160 861	0.169 057	0.156 762	0.140 369	0.138 320

序列名称	QCIF（176×144）		序列名称	CIF（354×288）
	5frame/s	15frame/s		30frame/s	15frame/s
	bitrate saving	bitrate saving		bitrate saving	bitrate saving
news_qcif	?0.01%	0.82%	news_cif	?0.21%	?0.41%
paris_qcif	2.11%	1.85%	paris_cif	2.08%	2.12%
football_qcif	0.40%	0.59%	football_cif	0.17%	1.87%
tempete_qcif	1.75%	1.15%	tempete_cif	1.88%	1.05%
bus_qcif	1.52%	0.97%	bus_cif	0.02%	0.42%
foreman_qcif	?0.32%	?0.80%	foreman_cif	?0.57%	?0.22%
mobile_qcif	2.22%	1.40%	mobile_cif	0.75%	1.26%