Reducing Mis-registration and Shadow Effects on Change Detection in Wetlands

Abstract

With respect to the inevitable mis-registration and shadow effects on change detection analysis, we propose objectbased post-classification of the Multivariate Alteration Detection components (OB-MAD). Very high spatial resolution images of drained, managed wetland ponds were used to compare the proposed OB-MAD method with three commonly used classification methods in terms of minimizing the influence of mis-registration and shadow on the change detection analysis: (a) the traditional MAD method with thresholds (Threshold-MAD), (b) a pixel-based postclassification of MAD components with decision tree analysis (PB-MAD), and (c) a traditional object-based postclassification method (OB-traditional). The OB-MAD method, which utilizes shape and textural information of objects derived from MAD components, produced the highest accuracy with respect to wetland change detection and successfully minimized the influence from the geometric distortion and shadow on the changed area.