RF classification black areas with more classes

hey guys
Trying to run an RF classification of a small scene of 10 classes. Below shows training data (kept quite small because there are some classes that are pixel limited)

As you can see below when i run classifier with all the classes and default RF settings, (tweaking parameters didn’t help) most of the image comes up as non classified:

This decreases with the inclusion of less classes:

What’s going wrong? Is this because RF can’t decide on the classification of these pixels? Is there a minimum sampling number for RF?

Any help greatly appreciated

The area apparently has less classes than 10, this could be one reason, also the distribution of training area could represents different pixels of the same class considered by you, the other issue from your screenshot, Did you re-sample the S2 image to one band resolution? , Did you subset the unwanted bands?

Take a look at this tutorial

Urban classification with S1

And this one

Deforestation monitoring with S1

Then please take a look at,

Crop mapping with Sentinel 2

And also this topic is important to you,

Classification based on spectral library

This thread is discussed in here,

Supervised and unsupervised classification

1 Like

thank you @falahfakhri :slight_smile: