The effect of correlation on the accuracy of the combined classifier in decision level fusion

Recently, decision level fusion has shown to have great potential to increase classification accuracy beyond the point reached by an individual classifier alone. There is a considerable body of literature about identifying optimal ways to combine classifiers; however, the selection of the classifiers to be combined is equally, if not more, crucial if an improvement is to be made. Agreement among classifiers can inhibit the gains obtained regardless of the method used to combine them. In this work the authors are assessing the level of agreement between different classifiers used in remote sensing using statistical measures. A study is performed in which an image is classified with several methods with different degrees of agreement between them. The results are then combined using decision fusion schemes and the increase of accuracy is observed.