Object Classification from a Hyper Spectral Image Using Spectrum Bands with Wavelength and Feature Set

Hyperspectral imaging is increasingly adopted for various remote sensing purposes with the increasing usage of sensors in various applications. One of the significant challenges in processing hyperspectral images is to make the data ready for carrying out an effective unmixing. However, this is yet an open-end problem as there is no apriori information to the system. Therefore, the proposed system introduces a simplified analytical model where an abundance-based methodology is used for extracting a useful feature followed by classification using a simple and non-iterative matrix-based operation. The proposed system emphasizes an unmixing process where the data is analyzed concerning data cube, wavelength, and feature set. Simulated in MATLAB, the study outcome shows that the proposed system offers higher classification accuracy and lower processing time.