Fuzzy clustering techniques for remote sensing images analysis

1. The dissertation proposes two unsupervised fuzzy c-means clustering algorithm (FCM), including DFCM and IFCM. DFCM algorithm proposes using density information for selecting initial centroids for FCM algorithm. IFCM algorithm proposes to using the spectral clustering and spatial information as a preprocessing step to map the original data space to a new space based on the main components. The proposed methods can improve the accuracy of the algorithm compared to the original algorithm.
2. The dissertation develops three semi-supervised fuzzy c-means clustering algorithms, including SMKFCM, SFCM-PSO and GIT2SPFCM-PSO. SMKFCM proposes the multiple-kernel technique to make data better separated; moreover, it uses labelled data to adjust the focus during clustering with the hope that the algorithm runs more stable. SFCM-PSO is a hybrid algorithm between semi-supervised method and PSO optimization technique. GIT2SPFCMPSO is a hybrid clustering algorithm developed by the semi-supervised possibilistic fuzzy c-means clustering based on interval type-2 fuzzy set with the parameters optimized by PSO technique. By using PSO technique for finding the optimal parameters. The proposed methods achieve better accuracy than existing methods.
The proposed methods can be applied to many types of RS images (radar, optics) and spatial resolutions (10m, 30m). Most of the experiments are used to the problem of the land cover classification of RS images. Although some limitations exist, the proposed methods can provide significantly better classification results than some recent other classification methods