分区域收敛的快速码书训练算法

庞朝阳; 孙世新; 潘晔; 龚海莹

分区域收敛的快速码书训练算法

计量
- 文章访问数: 2312
- HTML全文浏览量: 106
- PDF下载量: 687
- 被引次数: 0
出版历程
- 收稿日期: 2001-02-11
- 修回日期: 2001-10-10
- 刊出日期: 2002-09-19

A fast codebook training algorithm using local clustering

摘要

摘要: 该文提出分区域收敛的快速码书训练算法LC,它与LBC算法相比,结构简单、速度快,用典型的测试图像Lena和Barbara做实验,表明LC算法峰值信噪比只比LBG算法少2％左右,但运行速度东LBG的4.61-13.6倍。在比特率为0.375 bpp条件下,LC算法与LBG算法的重建图像质量无明显差别。
- 向量量化; LBG算法; LC算法; 峰值信噪比; 算法运行时间
Abstract: A fast codebook training algorithm using local clustering, named the LC algorithm, is presented in this paper. Comparing with the LBG algorithm, it is simple and fast. The experiment using typical test, images Lena and Barbara shows that the algorithm is 14.61 ~ 13.6 times faster than the LBG algorithm, with its PSNR being reduced by 2% approximately from the LBG algorithm only. The reconstructed images between the LC and the LBG algorithms have no obvious differences at the 0.375 bit/pixel condition.

HTML全文

参考文献(1)

Y. Linde, A. Buzo, R. M. Gray, An algorithm for vector quantization design, IEEE Trans. onCommum, 1980, COM-28(1), 84-95.[2]高文著,多媒体数据压缩技术,北京,电子工业出版社,1994年4月第一版,78-111.[3]P. Franti, T. Kaukoranta et al., Fast and memory efficient implementation of the exact PNN,IEEE Trans. on IP, 2000, 9(5), 773-777.[4]Chin-Chen Chang, Yu-Chen Hu, A fast LBG codebook training algorithm for vector quantization,IEEE Trans.on Consumer Electronics, 1998, 44(4), 1201-1208.[5]S. Pan, F. R. McInnes, M. A. Jack, Fast clustering algorithm for vector quantization, PatternRecognition, 1996, 29(3), 511-518.[6]T. Kaukoranta, P. Franti, O. Nevalainnem, Vector quantization by lazy pairwise nearest neighbormethod, Opt. Eng, 1999, 28(11), 1862-1868.[7]V. Ramasubramanian, K. Paliwal, Fast K-dimensional tree algorithms for nearest neighbor searchwith application to vector quantization, IEEE Trans. on SP, 1992, 40(3), 518-531.[8]C.K. Chan, L. M. Po, A Complexity reduction technique for image vector quantization, IEEETrans. on IP, 1992, 1(3), 312-321.[9]A. Buzo, A. H. Gray, R. M. Gray, J. D. Markel, Speech coding based upon vector quantization,IEEE Trans. on, ASSP, 1980, 28(5), 562-574.[10]张基宏,何振亚,一种指数型模糊学习矢量量化图像编码算法,通信学报,1998,19(10),1-6.[11]D. Comaniciu, Model-based training set synthesis for vector quantization, Proc. of the IASTEDInternational Conferrence on Signal and Image Processing, Nassau, Bahamas, October, 18-21,1999, 261-265.[12]D. Comaniciu, Trainng set synthesis for entropy-constrained transform vector quantization, Proc.IEEE ICASSP, Atlanda, 1996, 4, 2036-2039.[13]C-M. Huang, R. W. Harris, A comparison of several vector quantization codebook generationapproaches, IEEE, Trans. on IP, 1993, 2(1), 108-112.[14]沈兰荪著,图像编码与异步传输,北京,人民邮电出版社,1998年5月第1版,1998年5月第1次印刷,26-27.[15]M.R. Garey, D. S. Johnson, H. S. Witsenhausen, The complexity of the generalized Loyd-Maxproblem, IEEE Trans. on IT, 1982, 28(2), 255-256.[16]J. Shanbehzadeh, P. O. Ogunbona, On the computational complexity of the LBG and PNNalgorithms, IEEE Trans. on IP, 1997, 6(4), 614-616.[17]Bernd Fritzke, The LBG-U Method for Vector Quantization-An Improvement over LBG Iinspired from Neural Networks, Klawer Academic Publisher, 1997, 1-10.

施引文献

资源附件(0)

访问统计