Kernel Principle Component Analysis in Pixels Clustering
Jing Li, Dacheng Tao, Weiming Hu et al. · 2005
We propose two new methods in the nonlinear kernel feature space for pixel clustering based on the traditional KMeans and Gaussian mixture model (GMM). Unlike the previous work on the kernel machines, we give out a new perspective on the new developed kernel machines. That is, kernel principle component analysis (KPCA) combined with the KMeans and the GMM are kernel KMeans (KKMeans) and kernel GMM (KGMM), respectively. In this paper, we prove the new perspective on KKMeans and give out a clear statement on the KGMM as well. Based on this new perspectives, we can implement the KKMeans and the KGMM conveniently. At the end of the paper, we utilize these new algorithms on the problem of the colour image segmentation. Based on a series of experimental results on Corel colour images, we find that the KKMeans and KGMM can outperform the traditional KMeans and GMM consistently, respectively.
Cited by 15 works. doi:10.1109/wi.2005.86
Authors
References
- Pattern Classification
- Normalized cuts and image segmentation
- Pattern Classification (2nd ed.)
- Blobworld: image segmentation using expectation-maximization and its application to image querying
- Unsupervised segmentation of color-texture regions in images and video
- Edge flow: A framework of boundary detection and image segmentation
- Unsupervised multiresolution segmentation for images with low depth of field
- Estimation of generalized mixtures and its application in image segmentation
- Bayesian face recognition using Gabor features
- Cluster validation for unsupervised stochastic model-based image segmentation