A Wavelet-Based Hyper spectral Image Analysis to Characterize Soil Types

this paper introduces wavelet-based techniques for analyzing hyperspectral photos of soil to speedy and as it should be decide its land use or soil environment. Especially, it employs the wavelet border descriptors (WBD) method to calculate the hyper-spectral texture features of soil from hyperspectral photos. Those texture capabilities are then fused with the suggest spectrum extracted from the measured hyperspectral information to test various type models. Effects show that when the usage of the fusion approach, class accuracies improved as much as 4.6%. This study demonstrates that more than one-degree wavelet descriptors have potential for use in an automatic soil type classification machine for small-scale, local land-use mapping programs.