In molecular chemical imaging, specific recipes for material identification combine a collection of wavelength and algorithm parameters. These distinct recipes first require… a Chef.
Spectral Chef from ChemImage is wavelength analysis software using statistical analysis and machine learning techniques. It enables the discovery of differences between materials present in hyperspectral imagery based on spectral signatures.
The software enables users to select multiple Regions of Interest (ROI), view spectra, and select the combination of wavelengths that gives the best discrimination. Once the ROI is selected, the user can see the score image, threshold image, ROC curve (Receiver Operating Characteristic), and a histogram of in-class versus out-of-class values.
The results of this can be used to discriminate the selected materials from one another by collecting wavelengths selected in Spectral Chef that can be used in a recipe within Spectral Kitchen -- our landmark materials identification & classification software.