I'm trying to run a supervised classification on a landsat 8 (L8 OLI/TIRS COLLECTIONS LAND SURFACE REFLECTANCE ON-DEMAND) imagery. I created a signature file, then ran the Maximum Likelihood classification tool, but the output wasn't what I anticipated- which was to get a classified raster. Instead I got a black and white raster with high to low values ranging from 152-62.

I used the Image Classification Window (training samples, maximimum likihood, etc.) as seen: enter image description here

I created the training samples with the draw polygon option, then used the Training Sample Manager to create he classes. The classes were, Agriculture, Water, Urban, and Bare Earth. The values used were automatically generated by ArcMap, as seen below: enter image description here

The parameters used were as follows: enter image description here

The results were: enter image description here

I tried running an unsupervised classification, and that did s classification, so I'm not exactly sure what I'm doing wrong.

  • What type of classes are you working with? Did you select pixels on screen to generate training data? What values did you assign the classes in the training data?
    – Aaron
    Dec 31, 2018 at 4:00

1 Answer 1


Thank you @Aaron for your question, it forced me to re-examine the automated values that were generated for the training samples by ArcMap. I changed the values to consecutive (1-4) to represent the classes and it worked. My classifications were generated successfully! Key thing to remember when running the tool: Once training samples are created and merged into appropriate classes, assign values that would correspond to each class. Do not use the default figures.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.