Spatial-temporal Change Detection in Urban Area Using Adaptive Temporal Subset Multi-temporal InSAR Method

Abstract

Synthetic aperture radar (SAR) images are able to detect changes in an urban area with short revisited time. Most presented change detection methods based on SAR images are conducted using couples of the images. However, the change detection results are not reliable since the amplitude observations are sensitive to the change of surroundings and the types of changes are unknown. Here we propose a new change detection method using an adaptive temporal subset multi-temporal InSAR method. Single pixel change detection is developed using amplitude time series in order to identify the step-times: changes in the temporal domain. Then the parameters, e.g. deformation velocity are estimated and the coherent intervals are determined using interferometric phase time series. With the identified TCS, we distinguish different types of changes based on their coherent intervals. The main advantages of our method are reliable unsupervised change detection and detecting different types of changes without additional information. Experimental results by proposed method show that both appearing and disappearing buildings with their step-times are successfully identified and results by COSMO-Skyed ascending and descending images show a good agreement.