RS-MSSF Frame: Remote Sensing Image Classification Based on Extraction and Fusion of Multiple Spectral-Spatial Features.

Classifying remote sensing images with high spectral and spatial resolution became an important topic and challenging task in computer vision and remote sensing (RS) fields because of their huge dimensionality and computational complexity. Recently, many studies have already demonstrated the efficiency of employing spatial information where a combination of spectral and spatial information in a single classification framework have attracted special attention because of their capability to improve the classification accuracy. Shape and texture features are considered as two important types of spatial features in various applications of image processing. In this study, we extracted multiple features from spectral and spatial domains where we utilized texture and shape features, as well as spectral features, in order to obtain high classification accuracy. The spatial features considered in this study are produced by Gray Level Co-occurrence Matrix (GLCM) and Extended Multi-Attribute Profiles (EMAP), while, the extraction of deep spectral features is done by Stacked Sparse Autoencoders. The obtained spectral-spatial features are concatenated directly as a simple feature fusion and are fed into the Support Vector Machine (SVM) classifier. We tested the proposed method on hyperspectral (HS) and multispectral (MS) images where the experiments demonstrated significantly the efficiency of the proposed framework in comparison with some recent spectral-spatial classification methods and with different classification frameworks based on the used extractors.