SVM- and MRF-Based Method for Accurate Classification of Hyperspectral Images

TLDR

Hyperspectral sensors acquire many spectral bands, enabling finer material discrimination but posing new classification challenges. The study proposes a novel method for accurate spectral‑spatial classification of hyperspectral images. The method first applies a probabilistic support‑vector‑machine pixelwise classification, then refines the results with a Markov‑random‑field regularization that incorporates spatial context. The approach yields higher classification accuracy than recent advanced spectral‑spatial techniques on three airborne hyperspectral datasets.

Abstract

The high number of spectral bands acquired by hyperspectral sensors increases the capability to distinguish physical materials and objects, presenting new challenges to image analysis and classification. This letter presents a novel method for accurate spectral-spatial classification of hyperspectral images. The proposed technique consists of two steps. In the first step, a probabilistic support vector machine pixelwise classification of the hyperspectral image is applied. In the second step, spatial contextual information is used for refining the classification results obtained in the first step. This is achieved by means of a Markov random field regularization. Experimental results are presented for three hyperspectral airborne images and compared with those obtained by recently proposed advanced spectral-spatial classification techniques. The proposed method improves classification accuracies when compared to other classification approaches.