Change analysis of dual polarimetric Sentinel-1 SAR image time series using stationary wavelet transform and change detection matrix
Résumé
Change detection is a difficult task which intends to determine the changes of specific features within a certain time interval.
These change features are various and come from different causes. Synthetic aperture radar (SAR) imagery is particularly
appropriate for change detection thanks to the data acquisition possibility independently of Sun illumination and cloud coverage
of SAR system. The launch of new generation satellites allows to collect SAR image time series of fine spatial/temporal
resolution images from repeat-pass orbits. The Sentinel-1 SAR data are recently supplied free of charge for all data users, thus
opening new opportunities and challenges to process new data.
This paper provides initial change detection results on a time series of 10 dual polarimetric IW level-1 single look complex
(SLC) Sentinel-1 SAR images acquired in descending pass over Chamonix-Mont-Blanc, France. The presence of changes occurring
in the time series is identified in the change detection matrix (CDM) constructed by similarity cross tests between
wavelet extracted features issued from stationary wavelet transform (SWT).