Abstract
Abstract. The processing of hyperspectral remote sensing data, for information retrieval, is challenging due to its higher dimensionality. Machine learning based algorithms such as Support Vector Machine (SVM) is preferably applied to perform classification of high dimensionality data. A single-step unified framework is required which could decide the intrinsic dimensionality of data and achieve higher classification accuracy using SVM. This work present development of a SVM-based dimensionality reduction and classification (SVMDRC) framework for hyperspectral data. The proposed unified framework was tested at Los Tollos in Rodalquilar district of Spain, which have predominance of alunite, kaolinite, and illite minerals with sparse vegetation cover. Summer season image was utilized for implementing the proposed method. Modified broken stick rule (MBSR) was used to calculate the intrinsic dimensionality of HyMap data which automatically reduce the number of bands. Comparison of SVMDRC with SVM clearly suggests that SVM alone is inadequate in yielding better classification accuracies for minerals from hyperspectral data rather requires dimensionality reduction. Incorporation of modified broken stick method in SVMDRC framework positively influenced the feature separability and provided better classification accuracy. The mineral distribution map produced for the study area would be useful for refining the areas for mineral exploration.
Highlights
In the recent decade, advances in hyperspectral technology have increased the perception and knowledge of the earth’s surface
The objective has been achieved by the following sub-objectives (i) developing a Support Vector Machine (SVM) based unified framework for dimensionality reduction and classification of hyperspectral datasets (ii) finding the influence of the dimensionality reduction on the feature extraction (iii) comparison of classification accuracies derived from proposed approach vis-à-vis conventional approach of SVM classification
SVM-based dimensionality reduction and classification (SVMDRC) is the work which includes the developing of the framework and applying it on the hyperspectral image, and SVMC part is performing SVM classification on the hyperspectral image using the same training samples used by SVMDRC
Summary
Advances in hyperspectral technology have increased the perception and knowledge of the earth’s surface. The optimum number of bands are to be selected which have the total data content of the original image This is done by considering the virtual dimensionality of the reduced image. SVM as a dimensionality reducer and classifier was used for non-spatial dataset (Yang, 2009) that have lesser dimensions compared to the hyperspectral images. Motivated by this an algorithm is introduced that classifies an image along with dimensionality reduction, using SVM. The main goal is to develop a unified framework of SVM-based dimensionality reduction and classification algorithm for hyperspectral datasets and to evaluate its performance. The objective has been achieved by the following sub-objectives (i) developing a SVM based unified framework for dimensionality reduction and classification of hyperspectral datasets (ii) finding the influence of the dimensionality reduction on the feature extraction (iii) comparison of classification accuracies derived from proposed approach vis-à-vis conventional approach of SVM classification
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have