一种基于运动模式识别的视频对象编码算法

doi:10.13190/jbupt.200704.74.lihs

北京邮电大学学报 ›› 2007, Vol. 30 ›› Issue (4): 74-77.doi: 10.13190/jbupt.200704.74.lihs

一种基于运动模式识别的视频对象编码算法

黎洪松,许保华

(北京师范大学信息科学与技术学院, 北京 100875)

收稿日期:2006-09-05 修回日期:2007-03-28 出版日期:2007-08-30 发布日期:2007-08-30
通讯作者: 黎洪松

An Object-based Video Coding Using Motion Pattern Recognition

LI Hong－song,XU Bao－hua

(College of Information Science and Technology, Beijing Normal University, Beijing 100875, China)

Received:2006-09-05 Revised:2007-03-28 Online:2007-08-30 Published:2007-08-30
Contact: LI Hong－song

摘要/Abstract

摘要：

针对块匹配运动估计补偿（ME+MC）算法的不足，提出了一种基于自组织特征映射（SOM）的运动模式识别（MPR）算法，应用于会议电视的视频对象编码. 为了进一步提高SOM算法的性能，提出了频率敏感自组织特征映射（FSSOM）算法. 实验结果表明，FSSOM－MPR算法较ME+MC算法具有更好的编码性能，在压缩比为180∶1时，重建视频图像的平均峰值信噪比（PSNR）有3.3 dB的改善.

关键词: 运动估计补偿, 模式识别, 自组织特征映射, 视频对象编码

Abstract:

To improve the performance of block matching algorithm in motion estimation and motion complement (ME+MC)，a motion pattern recognition (MPR) algorithm based on selforganizing feature maps (SOM) is proposed and tested in object－based conference video coding. A frequency sensitive self－organizing feature maps (FSSOM) algorithm is proposed to improve the SOM algorithm’s performance. Experimental results show that FSSOM－MPR algorithm has better coding performances than ME+MC algorithm. When the compression ratio is 180∶1, the peak signal to noise ratio（PSNR） improvement of reconstructed video images is about 3.3 dB.

Key words: motion estimation and motion complement, motion pattern recognition, self-organizing feature maps, object-based video coding

中图分类号:

黎洪松,许保华. 一种基于运动模式识别的视频对象编码算法[J]. 北京邮电大学学报, 2007, 30(4): 74-77.

LI Hong－song,XU Bao－hua. An Object-based Video Coding Using Motion Pattern Recognition[J]. JOURNAL OF BEIJING UNIVERSITY OF POSTS AND TELECOM, 2007, 30(4): 74-77.

参考文献

[1] Hang H M, Chou Y M, Cheng S C. Motion estimation for video coding standards[J]. Journal of VLSI Signal Processing, 1997, 17(2/3): 113-136.
[2] 江涛, 张兆扬, 马然, 等. 一种将增强层运动补偿与FGST融合的视频编码方案[J]. 北京邮电大学学报, 2005, 28(2): 25-29. Jiang Tao, Zhang Zhaoyang, Ma Ran, et al. A video coding scheme combined FGST and motion compensation on enhancement-layer[J]. Journal of Beijing University of Posts and Telecommunications, 2005, 28(2): 25-29.
[3] Kohonen T. An introduction to neural computing[J]. Neural Networks, 1988, 1(1): 3-16.
[4] Seiffert U, Michaelis B. Estimating motion parameters with three-dimensional self-organizing maps[J]. Information Science, 1997, 101(3/4): 187-201.
[5] Cirrincione G, Cirrincione M. A novel self-organizing neural network for motion segmentation[J]. Applied Intelligence, 2003, 18(1): 27-35.
[6] Ebrahimi T, Horne C. MPEG-4 natural video coding—an overview[J]. Signal Processing: Image Communications, 2000, 15(4/5): 365-385.
[7] Shamim A, Robinson J A. Object-based video coding by global-to-local motion segmentation[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2002, 12(12): 1106-1116.
[8] Chaumonz M, Pateux S, Nicoias H. Object-based video coding using a dynamic coding approach//2004 International Conference on Image Processing (ICIP). Rennes: IRISA, Campus de Beaulieu, 2004: 1105-1108.
[9] Kim B C, Park R H. A fast automatic VOP generation using boundary block segmentation[J]. Real-Time Imaging, 2004, 10(2): 117-125.
[10] Wang Y, Osetermann J, Zhang Y Q. Video processing and communications[M]. Beijing: Publishing House of Electronics Industry, 2003: 245-246.
[11] Karayiannis N B, Li Y. A replenishment technique for low bit-rate video compression based on wavelets and vector quantization[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2001, 11(5): 658-663.

[1]	王智广, 王雷. 设计模式自动识别的研究进展[J]. 北京邮电大学学报, 2018, 41(4): 119-124.
[2]	王雷, 王智广. 考虑多种特征因素的设计模式自动识别[J]. 北京邮电大学学报, 2017, 40(5): 36-42.
[3]	刘渭滨, 邹智元, 邢薇薇. 模式分类中的特征融合方法[J]. 北京邮电大学学报, 2017, 40(4): 1-8.
[4]	古恒, 陈钊, 王枞, 张思悦, 傅群超. 双层聚类模型在日志数据分析中的应用[J]. 北京邮电大学学报, 2015, 38(s1): 63-66,71.
[5]	王景尧,宋梅,张勇，张英海. 认知网络中基于小波分析的切换判决[J]. 北京邮电大学学报, 2010, 33(6): 98-102.
[6]	高俊祥;杜海清;刘勇. 基于光照不变特征的椭球法运动阴影检测[J]. 北京邮电大学学报, 2009, 32(5): 109-113.
[7]	苏放范英磊. 基于Fuzzy模式识别的异构网络端到端丢包评判算法[J]. 北京邮电大学学报, 2007, 29(s1): 52-56.
[8]	陈光, 张洪刚, 郭军. 基于余弦整形变换的手写汉字训练样本生成方法[J]. 北京邮电大学学报, 2006, 29(s2): 118-121.
[9]	黎洪松, 刘洪伟. 新的学习矢量量化初始码书算法[J]. 北京邮电大学学报, 2006, 29(4): 33-35.
[10]	胡云伍宏涛张涵钰钮心忻杨义先. 大容量索引图像水印方案的设计与实现[J]. 北京邮电大学学报, 2005, 28(1): 26-29.
[11]	王波涛, 吴伟陵, 吴善培. 一种改进的神经网络训练算法[J]. 北京邮电大学学报, 2004, 27(4): 87-91.
[12]	邵旭, 丁炜, 石晶林, 张志群. 一种QoS自适应定量分类方法[J]. 北京邮电大学学报, 2001, 24(3): 60-64.
[13]	马晓敏, 杨义先. 基于内积方向的神经网络学习算法及应用^*[J]. 北京邮电大学学报, 1998, 21(4): 43-47.

一种基于运动模式识别的视频对象编码算法

An Object-based Video Coding Using Motion Pattern Recognition

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 13

编辑推荐

Metrics

本文评价