Supervised classification of EO data uses a set of samples from known patterns (usually reflectance spectra) to decide whether a given pattern belongs to one class or another. In landcover applications, one goes to the field, and observes that a given location is indeed class $\omega_c$. These ground observations usually ...

So I have just spent a week visiting at the VU Amsterdam. During this visit, I have also given a talk on EOLDAS. The slides are availabe from here.

In a previous post, I demonstrated how to stitch and put together a number of MODIS data files. This is useful and interesting, but in the end, we are interested in analysing the data we get out of the satellite. One first way around this might be to extract time ...