I am trying to imply a workflow on Sentinel-1 data for change detection (regarding changes in land use mostly) in an area with dense vegetation. I followed this tutorial http://step.esa.int/docs/tutorials/S1TBX%20TOPSAR%20Interferometry%20with%20Sentinel-1%20Tutorial_v2.pdf but I am not sure it is the right one, maybe it refers to land displacements. Can anyone suggest me in which direction I should search?
Thanks!
I think the interferometry tutorial is not quite the right one. Maybe you have a look at the functionalities of the Mask Manager in here: Time-series analysis with Sentinel-1
Also this one is interesting: http://www.un-spider.org/advisory-support/recommended-practices/mudslides-flood-sentinel-1/step-by-step
Depends a bit on what kind of change detection you want to apply - there are pre-classification methods where you directly compare the changes in backscatter, such as the ones above, and there are post-classification change detection methods, where you first classify landcover in two (or more images) and then compare the classified results.
thank you for your reply!I prefer pre-classification methods , I 'll chech your suggestions!
a tutorial on landcover classification is currently being prepared. But these two might also be helpful in terms of methods:
The tutorial on landcover classification was finally released: Landcover classification with Sentinel-1 GRD