A technique for feature selection in multiclass problems

One of the main phases in the development of a system for the classificatio n of remote sensing images is the definition of an effective set of feature s to be given as input to the classifier. In particular, it is often useful to reduce the number of features available, while saving the possibility t o discriminate among the different land-cover classes to be recognized. Thi s paper addresses this topic with reference to applications that involve mo re than two land-cover classes (multiclass problems). Several criteria prop osed in the remote sensing literature are considered and compared with one another and with the criterion presented by the authors. Such a criterion, unlike those usually adopted for multiclass problems, is related to an uppe r bound to the error probability of the Bayes classifier. As the objective of feature selection is generally to identify a reduced set of features tha t minimize the errors of the classifier, the aforementioned property is ver y important because it allows one to select features by taking into account their effects on classification errors. Experiments on two remote sensing datasets are described and discussed. These experiments confirm the effecti veness of the proposed criterion, which performs slightly better than all t he others considered in the paper. In addition, the results obtained provid e useful information about the behaviour of different classical criteria wh en applied in multiclass cases.