I have time series images and i want to perform a classification in a single time with all images.
In case you are using radar data you have to calibrate them first in order to remove acquisition-dependent radiometric distortions.
In a second step you use the coregistration tool to make one stack of your images. You can then use this stack as your feature space in the classification module.
I have to perform all the pre-processing tool (i.e. calibration, GLCM addition, Terrain correction) in all the images. Or simply i have to calibrate all the images and than i will perform coregistration tool??? Please help me.
First calibrate, then coregistrer, then calculate GLCM for selected bands, then create samples, then train and apply the classifier.