In this Getting Started section, we saw how to load scans from disk, define classes and build a pixel classifier.


The next step, in many applications, would be to define object segmentation to extract objects. This is useful to either classify objects or to extract further information from the objects (such as mean spectra, spectral indices or shape information)