Object-based Image Analysis (OBIA) segments an image by grouping pixels together into vector objects. Using these objects, you classify as land cover types.
Every object has its own unique composition. This is equivalent to it having its own spectral signature. Choose the spectral bands for ideal classification
Nearest neighbor classification is a powerful approach to classify segmented objects based on specific statistics and training samples chosen by the user.
Principal component analysis PCA identifies duplicate data over several datasets & aggregates essential information into groups called principal components.
We look at the digital image classification techniques in remote sensing (such as supervised, unsupervised & object-based) to extracts features of interest.