|
|
|
| Image texture clustering based on locality preserving projection |
| XING Rui1, ZHANG Yin1, ZHANG San-yuan1, ZHU Le-qing2 |
1. College of Computer Science and Technology, Zhejiang University, Hangzhou 310027, China;
2. College of Computer Science and Information Engineering, Zhejiang Gongshang University, Hangzhou 310018,China |
|
|
|
Abstract An improved image texture clustering method was proposed to solve the clustering problem in image retrieval. In feature extraction stage, dualtree complex wavelet transform（DTCWT）is applied to decompose image into tens of subbands. For those high frequency subbands, histogram signatures are generated as one of the texture features. In clustering stage, the distances between the data points are computed adaptively according to the data distribution density. Locality preserving projection is then employed on the distances to reduce the dimensionality of the data space. kmeans is used to cluster the data in the lower dimensionality space. Histogram signature can represent image texture well in multiple directions of the DTCWT decomposition. Moreover, the distance matrix built on data density can detect dataset locality effectively in combination with locality preserving projection. The experimental results show that the proposed method outperforms the traditional methods．
|
|
Published: 01 September 2010
|
|
|
基于保局映射的图像纹理聚类
为了解决图像检索中的聚类问题,提出一种改进的图像纹理聚类算法.在纹理特征提取阶段,采用双树复小波对图像进行分解,然后对每个高频段提取直方图签名作为纹理特征;在聚类阶段,根据数据分布的密度来动态地计算数据点的邻接矩阵,再采用保局映射进行降维,对降维后的数据进行kmeans聚类.通过采用直方图签名的方式能有效地表示图像纹理在各个方向上特征信息,同时根据数据密度构建的邻接矩阵,能够和保局映射一起更有效地发掘数据之间的局部相关性.实验表明：相对于传统方法,该算法具有更高的聚类正确性.
|
|
| ［1］ RUI Y, HUANG T S. Image retrieval: current techniques, promising directions and open issues ［J］. Journal of Visual Communication and Image Representation, 1999（10）:3962．
［2］ SMEULDERS A W M, WORRING M, SANTINI S, et al. Contentbased image retrieval at the end of the early years ［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2000, 22(12):13491380．
［3］ KINGSBURY N. Complex wavelets for shift invariant analysis and filtering of signals ［J］. Journal Applied and Computational Harmonic Analysis, 2001, 10(3):234253.
［4］ JOLLIFFE I T. Principal component analysis ［M］. New York: SpringerVerlag, 1989: 150165.
［5］ DUDA R O, HART P E, STORK D G. Pattern classification ［M］. 2nd ed. New Jersey: Wiley Interscience,2000: 216268．
［6］ WOUWER G V, SCHEUNDERS P, DYCK D V. Statistical texture characterization from discrete wavelet representation ［J］. IEEE Transactions on Image Processing, 1999, 8(4):592598．
［7］ CAI D, HE X. HAN J. Document clustering using locality preserving indexing ［J］. IEEE Transactions on Knowledge and Data Engineering, 2005, 17(12):16241637．
［8］ XU W, LIU X, GONG Y. Document clustering based on nonnegative matrix factorization ［C］∥Proceedings of the 26th Annual International ACM SIGIR Conference on Research and Development in Informaion Retrieval. Toronto:ACM, 2003:267273.
［9］ LOVASZ L, PLUMMER M. Matching theory ［M］. North Holland, Budapest: Akademiai Kiado, 1986:255306.
［10］ LAZEBNIK S, SCHMID C, PONCE J. A sparse texture representation using local affine regions ［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2005, 27(8):12651278． |
|
Viewed |
|
|
|
Full text
|
|
|
|
|
Abstract
|
|
|
|
|
Cited |
|
|
|
|
| |
Shared |
|
|
|
|
| |
Discussed |
|
|
|
|
