Spectral Divergence

Features:

Generates comparison between any two regions of interest interactively selected on the image using spectral information divergence (SID) algorithm.

Steps:

1.       Load the file. Select Spectra Mathematics and select Spectral Divergence.

2.       Click Select 1st spectrum. Draw a region of interest. A spectrum corresponding to the average spectrum of the selected region will appear in the SPECTRAL ANALYSIS panel.

3.       Click Select 2nd spectrum. Draw a region of interest. A spectrum corresponding to the average spectrum of the selected region will appear in the SPECTRAL ANALYSIS panel.

4.       Press Generate to measure the spectral information divergence score between two objects. A lower score corresponds to a stronger similarity between the areas. A value of 0 indicates no divergence between the two areas.

Press Clear ROI/Spectra to start another analysis or to perform any other spectral tasks.

 

Additional Information:

The method measures the spectral similarity between the two mean spectra from the selected ROIs by using the spectral information divergence (SID) algorithm. This method might be also used to compare the spectral signature of an unknown material against the reference area. The smaller the divergence, the more likely the pixels are similar.

The method computes spectral similarity scores based on the divergence between the probability distributions of the two spectra.

References:

Chein-I Chang. “An Information-Theoretic Approach to Spectral Variability, Similarity, and Discrimination for Hyperspectral Image Analysis.” IEEE Transactions on Information Theory 46, no. 5 (August 2000): 1927–32. https://doi.org/10.1109/18.857802.

Du, H., C.-I. Chang, H. Ren, F. M. D’Amico, and J. O. Jensen, J. "New Hyperspectral Discrimination Measure for Spectral Characterization." Optical Engineering 43, No. 8 (2004): 1777-1786.