Training and Visualizing your Data

In the Train tab, you can create your classes and assign some samples in the image to the classes. GeoDMA uses this information to train the supervised classification algorithm.

After creating your classes, you should select samples in your data. You can select one polygon per click, or more than one, defining a region, using buttons Pointer Cursor or Graphic Cursor, respectively.

When you select the samples from one class, go to GeoDMA window, select the class and click Associate. In TerraView, unselect all samples and perform the same for all your classes.

In the tab Visualization you can view scatterplots of your attributes, also showing the samples that you have selected in the Training step. Select the x and y axis, adjust the sampling rate and click Plot.

That's it! Now you have trained your supervised algorithm, and visualized your attributes!

Go back to the manual.