Hyperspectral Image Classification Based on the Gabor Feature with Correlation Information

Gabor filter is widely used to extract spatial texture features of hyperspectral images (HSI) for HSI classification; however, a single cannot obtain the complete image features. In paper, we propose an classification method that combines (GF) and domain-transformation standard convolution (DTNC) filter. First, use from first two principal components dimensionality-reduction with PCA. Second, DTNC correlation in all bands. Finally, Large Margin Distribution Machine (LDM) uses linear fusion kinds classify HSI. The experimental results show accuracy Indian Pines, Pavia University, Kennedy Space Center data sets 96.64, 98.23, 98.95% only 4, 3, 6% training samples, respectively; these accuracies are 2–20% higher than other tested methods. Compared information based on SVM, EPF, IFRF, PCA-EPFs, LDM-FL, GFDN method, proposed GFDTNCLDM, significantly improves classification.