Affiliation:
1. Hunan University of Technology
Abstract
Kernel methods are famous for their efficiency and robustness in processing non-linear machine learning problems in the high dimensional feature space, and thus widely applied in image classification and detection. The proper principal components are selected for KPCA reconstruction according to noise features. Finally, the improved image is obtained by performing inverse method. Experimental results show that the proposed method can suppress noise interference in remote sensing images, and preserve the useful information of original data more completely.
Publisher
Trans Tech Publications, Ltd.
Reference5 articles.
1. .
2. Brigger P, Hoeg J, Unser M.B-spline snakes: a flexible tool for parametric contour detection[R].IEEE Transactions on Image Processing, 2000, 9(9): 1484-1496.
3. Yue Wang and Eam Khwang Teoh.Dynamic B-snake model for complex objects segmentation[J].Image and Vision Computing, Volume 23, Issue 12, 1 November 2005, Pages 1029-1040.
4. Sun L and Luo J S. 2009. Three-dimensional hybrid denoising algorithm in derivative domain for hyperspectral remote sensing imagery. Spectroscopy and Spectral Analysis, 29(10): 2717–2720.
5. Vaiphasa C. 2006. Consideration of smoothing techniques for hyperspectral remote sensing. ISPRS Journal of Photogrammetry and Remote Sensing, 60(2): 91–99.
Cited by
1 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献