Dimensionality reduction and hierarchical clustering in framework for hyperspectral image segmentation

Abstract

The hyperspectral data contains hundreds of narrows bands representing the same scene on earth, with each pixel has a continuous reflectance spectrum. The first attempts to analysehyperspectral images were based on techniques that were developed for multispectral images by randomly selecting few spectral channels, usually less than seven. This random selection of bands degrades the performance of segmentation algorithm on hyperspectraldatain terms of accuracies. In this paper, a new framework is designed for the analysis of hyperspectral image by taking the information from all the data channels with dimensionality reduction method using subset selection and hierarchical clustering. A methodology based on subset construction is used for selecting k informative bands from d bands dataset. In this selection, similarity metrics such as Average Pixel Intensity [API], Histogram Similarity [HS], Mutual Information [MI] and Correlation Similarity [CS] are used to create k distinct subsets and from each subset, a single band is selected. The informative bands which are selected are merged into a single image using hierarchical fusion technique. After getting fused image, Hierarchical clustering algorithm is used for segmentation of image. The qualitative and quantitative analysis shows that CS similarity metric in dimensionality reduction algorithm gets high quality segmented image.

Clustering using kernel entropy principal component analysis and variable kernel estimator

“…As a solution, we consider the reduction of the dimensionality as an efficient preprocessing. Due to this effect, many researches have been done to get rid of this complication [18][19][20][21]. Thus, we establish a nonlinear method named KEPCA, which improves the existed linear EPCA method [13].…”

Integrated NIR-HE based SPOT-5 image enhancement method for features preservation and edge detection

“…Therefore, the NIR channel was used to refine the image's contrast. The HE method enhances the image by distributing the image brightness levels equally across the brightness scale [1], [11], [15], [21], [44], [53], [54], [72], [94], [119], [136], [140], [145], [150], [159], [168], [169], [172], [175]. Furthermore, the intensity of the contrast enhancement method is measured through the root mean square (RMS), where the higher the RMS value, the better the contrast image [22], [35], [48], [178]- [181].…”

“…Therefore, multispectral images store multiple values for each pixel, captured through the amount of light in different channels of the electromagnetic spectrum. The common multispectral images are RGB, which contain three channels that correspond to the R, G and B regions of the spectrum [39], [60], [84], [88], [94]. However, throughout the study, the false colour image is used, where the three channels correspond to the NIR, R and G regions.…”