7.2. Algorithmic approaches

An important step in image analysis is the interpretation of pixels as belonging to different land cover classes. This is generally termed as image classification. Useful information such as the spatial distribution and extent of different crops can only be derived after algorithms have been applied to satellite images. Different categories of algorithms exists for achieving this purpose. This section discusses the different approaches and algorithms applied to RS for such information retrieval. First, the two groups of classification techniques of supervised and unsupervised classification are described. Next, parametric and non-parametric classification algorithms are discussed. Finally, three common classification approaches are highlighted.