A Hybrid Classification Scheme for Mining Multisource Geospatial Data

Supervised learning methods such as Maximum Likeli- hood (ML) are often used in land cover (thematic) clas- sification of remote sensing imagery. ML classifier relies exclusively on spectral characteristics of thematic classes whose statistical distributions are often overlapping. The spectral response distributions of thematic classes are de- pendent on many factors including elevation, soil types, and atmospheric conditions present at the time of data acqui- sition. A second problem with statistical classifiers is the requirement of large number of accurate training samples, which are often costly and time consuming to acquire over large geographic regions. With the increasing availability of geospatial databases, it is possible to exploit the knowl- edge derived from these ancillary datasets to improve clas- sification accuracies even when the class distributions are highly overlapping. Likewise newer semi-supervised tech- niques can be adopted to improve the parameter estimates of statistical model by utilizing a large number of easily available unlabeled training samples. Unfortunately there is no convenient multivariate statistical model that can be employed for mulitsource geospatial databases. In this pa- per we present a hybrid semi-supervised learning algorithm that effectively exploits freely available unlabeled training samples from multispectral remote sensing images and also incorporates ancillary geospatial databases. We have con- ducted several experiments on real datasets, and our new hybrid approach shows over 15% improvement in classif- ciation accuracy over conventional classification schemes. Key Words: MLC, EM, Semi-supervised Learning.