X
PrevPrev Go to previous topic
NextNext Go to next topic
Last Post 05 Dec 2016 11:42 AM by  anon
Maximum Likelihood Classification Using Spectral Library Files
 2 Replies
Sort:
You are not authorized to post a reply.
Author Messages

anon



New Member


Posts:14
New Member


--
05 Dec 2016 11:42 AM
    Hi everyone, I have been trying to use endmember spectral I collected and saved as a spectral library file to classify surface features between different datasets. As such, ROIs are inconvenient to redraw and use for each new image. It seems as if the Maximum Likelihood Classification only works with ROI files, and not with spectral library files. Is this the case, or am I doing something wrong? Thanks in advance for your help.

    MariM



    Veteran Member


    Posts:2396
    Veteran Member


    --
    05 Dec 2016 12:00 PM
    You cannot use Maximum likelihood, mahalanobis distance or similar methods that require the calculation of covariance and matrix inversion. This is because a spectrum,whether collected from a single pixel or calculated from a group, is a single mean value. To calculate covariance requires multiple samples per training class so it is only designed to work with ROIs which typically comprise spectra from multiple pixels. I would try using a method such as SAM which does not require the calculation of covariance. If your images have spatial reference, you should be able to restore ROIs on each image - assuming they cover the same spatial extents.

    Deleted User



    New Member


    Posts:14
    New Member


    --
    06 Dec 2016 09:45 AM
    Thanks Mari. You are very helpful, I really appreciate your insight. I think that restoring ROI could work between datasets I have from the same scene. Much thanks, again! Alex
    You are not authorized to post a reply.