Wavelet image compression based on support vector machines
-
摘要: 提出了一种结合支持向量机(SVM,Support Vector Machines)回归与小波变换的新的静态图像压缩方法.SVM回归方法可以学习原始数据之间的相关性,并采用小部分训练样本,即支持向量来稀疏表示原始数据集,利用这一特性来逼近和约减小波系数,可以达到数据压缩的效果.首先采用小波变换把原始图像分解成不同尺度的多个子带,由于最低频子带系数非常重要,采用DPCM直接编码,然后对其它频带系数采用SVM回归进行压缩.由于不同尺度和方向的小波系数特征不同,为尽可能去除小波系数间的各种相关性,给出了适合SVM回归的小波系数的有效组织方式.最后研究了支持向量及其相应权重的混合编码方法.实验结果表明:与同类压缩方法相比,本算法获得的恢复图像的主客观质量有明显提高.Abstract: A novel image compression algorithm that combined SVM(support vector machines) regression and wavelet transform was presented. SVM regression could learn dependency from training data and realized compression by using fewer training point (support vectors) to represent the original data. Thus, wavelet coefficients could be compressed based on this feature. Image was decomposed into subbands of different scales in using wavelet transform. The lowest subband was coded using DPCM for its great importance, and the other coarser subbands were compressed by SVM. Since the characteristic of the wavelet coefficients was various in different scales and directions, it was a key problem to design the appropriate organization method of the coefficients. Effective entropy coding technique was also studied to encode the support vectors and the corresponding weights. Experiment results demonstrate the coding efficiency of the proposed algorithm.
-
Key words:
- image compression /
- support vector machines /
- wavelet transforms /
- entropy coding
-
[1] Namphol A, Chin S H, Arozullah M. Image compression with a hierarchical neural network[J]. IEEE Transactions on Aerospace and Electronic Systems, 1996, 32(1):326-338 [2] Chen O T C, Sheu B J, Fang W C. Image compression using self-organization networks[J]. IEEE Transactions on Circuits and Systems for Video Technology, 1994,4(5):480-498 [3] Amerijckx C, Verleysen M, Thissen P, et al. Image compression by self:organized Kohonen map[J]. IEEE Transactions on Neural Networks, 1998,9(3):503-507 [4] Vapnik V N. The nature of statistical learning theory[M]. New York:Springer-Verlag, 1995 [5] Robinson J, Kecman V. Combing support vector machine learning with the discrete cosine transform in image compression[J]. IEEE Transactions on Neural Networks, 2003, 14(4):950-958 [6] Smola A J, Scholkopf B. A tutorial on support vector regression .NeuroCOLT Tech,1998-030, 1998 [7] Drucker H, Burges C J C, Kaufmann L, et al. Support vector regression machines[M]. Cambridge,MA:MIT Press, 1997 [8] Vapnik V, Golowich S, Smola A J. Support vector method for function approximation, regression estimation, and signal processing Mozer M, Jor-dan M, Petsche T.Advances in Neural Information Processing Systems 9. Cambridge, MA:MIT Press,1997 [9] Mukherjee S, Osuna E, Girosi F. Nonlinear prediction of chaotic time series using support vector machines Proceedings of IEEE NNSP ’97.Amelia Island:FL,1997 [10] Antonini M, Barlaud M, Mathieu P, et al. Image coding using wavelet transform[J]. IEEE Transactions on Image Processing, 1992, 1(2):205-220 [11] Mallat S G. A theory for multiresolution signal decomposition:the wavelet representation [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1989, 11(7):674-693 [12] 汪雪林,韩华,彭思龙.基于小波域局部高斯模型的图像复原[J].软件学报,2004,15(3):433-450 Wang X L, Han H, Peng S L. Image restoration based on wavelet-domain local Guassian model [J]. Journal of Software, 2004, 15(3):433-450(in Chinese) [13] Chang Chih-Chung, Lin Chih-Jen. LIBSVM:a library for support vector machines . http://www.csie.ntu.edu.tw/cjlin/libsvm,2001
点击查看大图
计量
- 文章访问数: 2878
- HTML全文浏览量: 185
- PDF下载量: 1095
- 被引次数: 0