Adaptive Coherence Estimator (ACE) is derived from the Generalized Likelihood Ratio (GLR) approach. The ACE is invariant to relative scaling of input spectra and has a Constant False Alarm Rate (CFAR) with respect to such scaling. Similar to Constrained Energy Minimization (CEM) and Matched Filtering (MF), ACE does not require knowledge of all the endmembers within an image scene.

You can also write a script to perform ACE target detection using the SpectralAdaptiveCoherenceEstimator task.

  1. From the Toolbox, select Classification > Supervised Classification > Adaptive Coherence Estimator Classification. The Adaptive Coherence Estimator Input File dialog appears.
  2. Select an input file and perform optional spatial and spectral subsetting, then click OK. The Endmember Collection:ACE dialog appears.
  3. Import spectra to match. For details, see Endmember Options and Manage Endmember Spectra.
  4. Click Apply. The Adaptive Coherence Estimator Parameters dialog appears.
  5. Use the toggle button to select Compute New Covariance Stats and enter an output statistics filename, or toggle to Use Existing Stats File.
  6. If you selected Compute New Covariance Stats: To remove anomalous pixels before calculating background statistics, enable the Subspace Background check box. Then, specify in the Background Threshold field the fraction of the background in the anomalous image to use for calculating the subspace background statistics. The threshold range is 0.500 to 1.000 (the entire image).
  7. Select output to File or Memory.
  8. Click OK.
  9. If you selected Use Existing Stats File, select the statistics file that corresponds to the input data file when the Input File dialog appears. This statistics file must contain both the mean and covariance statistics for the input data.

Adaptive Coherence Estimator Results

The results of ACE appear as a series of gray scale images, one for each selected endmember.

The default stretch setting provides good visibility for small features. If needed, you can apply a different stretch so that larger features in the image are visible.

Note: You can set a default stretch range so that you do not have to stretch the data each time they are displayed.