Road Change Detection Using Object-oriented Projective Interactive Partition

Abstract

In this paper the authors present an object-oriented change detection method which uses projective interactive partition to detect the change of the urban road.The QuickBird images acquired in different years were used in the experiment,on which multi-scale segmentation and hierarchical classification were carried out for urban road extraction according to the spectrum,shape and texture features.Then the object layers were established in the same detection,and they were the projection layers taking over the classification results of asynchrony images.Interactive partition was also realized on these layers.Finally the change detection results were achieved after judging the consistency of land feature categories.The experiment results show that the method of object-oriented change detection using projective interactive partition can extract and detect the urban road effectively.