ABSTRACT
This chapter focuses on the extraction of hyperspectral information. The extraction of spectra from the hypercube relies on the identification of regions of interest (ROI). One way to identify ROIs is to take advantage of commercial software available. The other way to identify ROIs is to implement image processing algorithms, where segmentation of a two-dimensional image of the object is required. Hyperspectral imaging has shown its great power in wide applications particularly in the detection of food quality and safety. The data acquired using hyperspectral imaging systems are three dimensional with two dimensions for spatial coordinates and one dimension for wavelength scale. However, a hyperspectral image consists of many two-dimensional images at different wavelengths, it is therefore important to allocate appropriate bands for band math operation from the massive data space. Depending on the imaging modalities, the acquired hyperspectral images could be recorded in reflectance, absorbance or transmittance.
