The SAM Target Finder with BandMax Wizard guides you through a step-by-step process to find targets in hyperspectral images using a SAM classification and the BandMax algorithm. The BandMax part of the Wizard increases classification accuracy by determining an optimal subset of bands to help you separate your targets from known background materials.

You can also directly access the BandMax algorithm from the Endmember Collection dialog.

The BandMax algorithm was developed by the Galileo Group, Inc. The process is based on the United States patent application titled Spectral image processing system and method for target detection and identification. This document was submitted by Timothy J. Pachter and Daniel Matthew Puchalski to the United States Patent Office in 2002.

To start the wizard, select Target Detection > SAM Target Finder with BandMax from the Toolbox.

The SAM Target Finder with BandMax Wizard guides you through the following steps:

  1. Select Input Image: Select the input file and the root name of the output files for the Wizard.
  2. Select Targets: Select the spectra to use as targets.
  3. Select Background: Select backgrounds to suppress.
  4. Select Band Subset: Identify significant bands to use in a SAM analysis.
  5. Select SAM Parameters: Define a SAM maximum angle parameter.
  6. Examine SAM Results: Analyze and examine SAM results. You may exit the Wizard at this time if the results are acceptable.

The overall workflow of the Wizard is shown in the following figure.

If examining the results from this mapping process shows that the classification accuracy is not sufficient, the Wizard returns you to step 2, where you can refine your targets and backgrounds before running BandMax and the SAM analysis again. You may continue to refine targets, select backgrounds, use BandMax to find the appropriate band subset, and re-classify your input data with SAM. When your analysis is complete, you can save and view a report of the processing. ENVI adds the resulting output to the Layer Manager. As with any image processing tool, poor results may occur if you provide poor input data or inappropriate parameters.