Hi everyone,
I create a classification on planetscope data. I made the classification using three methods: random trees, maximum likelihood and support vector machine. Below screen with random forest classification.
![](/legacyfs/online/490722_rfclassific.png)
now I want to check the accuracy of the classification. I need a reference raster for this and I don't know how to do it well. Can someone explain it to me? I know the tool for creating accuary assessment points but I don't know what the reference raster should be.
I would also like to ask if my training samples are ok, I have photointerpretation key and orthophoto and I created samples based on this data. Is this amount of samples good?
![](/legacyfs/online/490723_trainsamples.png)