Spectral Signature Matching Toolbox
Note: Requires spectra saved in the ECOSTRESS format (name_spectrum.txt).
Features:
Compares the similarity of each pixel spectrum to the reference spectrum.
Detects a target in the hyperspectral image by using a specified spectral signature matching method.
Shows the area that corresponds to the reference spectrum in the ECOSTRESS format (https://www.mathworks.com/help/images/ref/readecostresssig.html). Spectra can be recorded in the ECOSTRESS format from a SPECTRAL ANALYSIS panel and from ENDMEMBERS EXTRACTION toolbox (see corresponding sections).
Calculates the area (in pixels) that matches a specific spectrum with the user input threshold value.
Steps:
1. Browse your directory to select the reference spectrum name_spectrum.txt file in the ECOSTRESS format.
2. Click Load Library Spectrum.
3. Select the method of spectral matching. The methods are described below. Available options are:
Spectral Angular Mapper (SAM)
Jeffries Matusita SAM (JM SAM)
Normalized Spectral Similarity (NS3)
Spectral Information Divergence (SID)
Spectral Information Divergence - Spectral Angular Mapper Hybrid (SIDSAM)
4. Click Match to visualize the matching score. A lower score indicates a better match with the uploaded spectrum.
5. The software automatically generates histograms of all score intensities under the HISTOGRAM OF SCORE MAP. Select a threshold close to the first feature (i.e., a peak) on the histogram.
6. The orange color object(s) are generated and overlapped with the original image. The total area of the match in pixels appears in the title of the IMAGE DISPLAY panel.
Additional Information:
Spectral Angle Mapper (SAM) is a spectral image processing classification that matches the spectra from pixels to reference spectra. The algorithm determines the spectral similarity between two spectra by calculating the angle α between the spectra. Smaller angles represent closer matches to the reference spectrum. Reference spectra can be uploaded from known sources or directly extracted from an image as region of interest (ROI) mean spectra (from the Main Interface see Main Interface manual).
JMSAM is a spectral image processing classification that matches the spectra from each pixel (test spectrum) to a reference spectrum. The method computes spectral similarity based on the Jeffries Matusita (JM) and SAM distances between two spectra. A smaller JMSAM score indicates a strong match between the test spectrum and the reference spectrum.
NS3 computes spectral similarity based on the Euclidean and SAM distances between the test spectrum and the specified reference spectrum by using the normalized spectral similarity score (NS3) method.
SID measures the spectral similarity between the spectra of each pixel in the hyperspectral data and the specified reference by using the spectral information divergence (SID) technique. The method computes spectral similarity based on the divergence between the probability distributions of the two spectra.
SIDSAM measures the spectral similarity between the spectra of each pixel in the hyperspectral data and the specified reference by using a hybrid SID-SAM hybrid score.
References:
SAM: Kruse, F. A., A. B. Lefkoff, J. B. Boardman, K. B. Heidebrecht, A. T. Shapiro, P. J. Barloon, and A. F. H. Goetz. "The Spectral Image Processing System (SIPS) - Interactive Visualization and Analysis of Imaging spectrometer Data." Remote Sensing of Environment 44 (1993): 145-163.
JMSAM: Padma, S., and S. Sanjeevi. “Jeffries Matusita Based Mixed-Measure for Improved Spectral Matching in Hyperspectral Image Analysis.” International Journal of Applied Earth Observation and Geoinformation 32 (October 2014): 138–51. https://doi.org/10.1016/j.jag.2014.04.001.
NS3: Nidamanuri, Rama Rao, and Bernd Zbell. “Normalized Spectral Similarity Score (NS3) as an Efficient Spectral Library Searching Method for Hyperspectral Image Classification.” IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 4, no. 1 (March 2011): 226–40. https://doi.org/10.1109/JSTARS.2010.208643
SID: 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.
SIDSAM: Chang, Chein-I. “New Hyperspectral Discrimination Measure for Spectral Characterization.” Optical Engineering 43, no. 8 (August 1, 2004): 1777. https://doi.org/10.1117/1.1766301.
