Abstract
This paper proposes the classification framework based on the Bayesian theory with the single polarization multi-temporal synthetic aperture radar (SAR) and an optical data, and incorporates the proposed training sample selection (SS) methods. Within this framework, the combination with gray level co-occurrence matrix (GLCM)-based textural measures is investigated. The two procedures of the classification and proposed SS are united, where SS generates the accurate and dispersed training samples. Extracted features from multi-temporal SAR data—namely, the average backscattering coefficient, the backscatter temporal variability, and the long-term coherence and the reflectance values from optical data, are integrated with the GLCM textural data. Classification results were generated by taking Osaka City, Japan, as the study area. The selected major classes were water bodies, woodlands, fields, and built-up areas. The most suitable data used for classification was the multi-temporal SAR and an optical data combination with the mean textural, because of the supplement of different data and the smoothing effect of the texture. The higher quality training samples obtained by using the combined Support Vector Machine (SVM) and Neural Network (NN)-based SS method for training the Bayesian classifier generated the highest classification accuracies in all of tested cases.
