Optimal Band Selection and Scale based Feature Selection for Hyper Spectral Image Classification using Hybrid Neural Network

The use of hyper spectral remote sensing to categorize images has been a popular study topic. Non-linear features and high dimensionality are common in Hyper Spectral Image classifications. Band selection can be used to cut computation costs and speed up knowledge discovery. In hyperspectral photographs, mixed pixels often include some ambiguity. This paper suggests a new band selection procedure method to address these issues. Band selection will be a useful tool for lowering the size of hyperspectral data while also assisting us in overcoming dimensionality issues. To automate hyperspectral picture analysis, this article uses a hybrid model. For ground/landcover classification using hyperspectral pictures, this suggested approach employs strategic and competitive theory models. In a classifier-ensemble system, there are also game theory application models for hyperspectral band grouping and pixel classification.