pixelcooker

How to remove clouds cover in Sentinel-2 Imagery using ArcGIS Pro's Raster Functions and Scene Classification Layer available in Level-2A Sentinel Imagery. 

This is the newest satellite imagery cloud removal and correction tutorial I have made. The workflow described in this tutorial is similar to my previous tutorial about satellite imagery clouds removal and correction using ArcGIS Desktop. The only difference is, the workflow I described in this video is better. Because this time, we could use the QA Band as the reference cloud mask. So, it is more simple, robust, accurate, precise, and guarantees a better result. It is all thanks to Graphical Raster Functions Editor implemented in ArcGIS Pro. 

Cloud cover is a long-time problem for optical imagery analysis in remote sensing. It is obscuring important earth surface object that became the main target on image analysis. There are numerous methods and efforts have been developed to fix the cloud cover problem. The solutions are coming from simple approaches like masking and replacing using a raster calculator or map algebra, to more complex ways like multitemporal pixels blending, fast Fourier transform-based filtering, or even machine learning methods.  The correction results also vary. Sometimes certain method could give a nearly perfect cloudless image, but sometimes it doesn't. One simple way to remove cloud in optical imagery is using the masking and replacing method. This method needs at least two imagery covering the same area at different dates, so it is expected that two imagery will have different cloud cover conditions, and finally, both of them can be merged together to create a cloudless image.  This approa...