Multi-temporal Remote Sensing Image Change Detection Based On Curve Evolution Model

Change detection using multi-temporal remote sensing images can provide a consistent view of land cover changes and ubran development,and is very important and significant.As a basic part of remote sensing technology,change detection has attracted many scholars to research.Nowadays,a large number of change detection methods and models have been proposed for different research applications from the perspective of pixel-based and object-based.In this paper,we try to modify active contour models to improve the accuracy of change detection.The main contents are as follows:(1)An automatic method for change detection in remote sensing images using level set evolution with neighborhood constraints is proposed.This method introduces the two-phase segmentation model with curve evolution theory into the change detection in order to improve the detection accuracy.Furthermore,to increase the robustness of the method against noise,we added neighborhood constraints to the levelset model.Our method not only considered the strategy of the level set initialization but also adopted a coarse-to-fine procedure.After implementing the proposed method in a multi-resolution framework,we validated the algorithm by numerical results of real remote sensing images.(3)A remote sensing image change-detection method based on level set evolution and support vector machine(SVM)classification is proposed.This method combines the pixel-based detection method with the object-based detection method and can select the appropriate SVM classifier training samples.In addition,it can effectively improve detection accuracy and the automation level of the algorithm.The experimental results show that the method based on the combination of level set evolution and SVM classification can greatly reduce the meaningless speckle noise,and the final detection result is more accurate.