Reclasify Deep learning model results (Land Cover Classification (Sentinel-2)).

Habal

I need to be help regarding the use of a Deep Learning model developed by ESRI for Land Cover Classification (Sentinel2).

Following the procedure dedicated to land cover classification with Sentinel (link:  https://www.arcgis.com/home/item.html?id=afd124844ba84da69c2c533d4af10a58 ), the ESRI-developed learning model worked very well.

However, the model used generates a classification into 15 classes. In my case, I would like to synthesize these classes into only 5 classes. The described procedure does not mention how to reduce the number of classes. I have tried various tools available in ArcGIS Pro, but to no avail. Therefore, I would appreciate your guidance on this matter.

NB: In the classification tool, none of the tools work.

JayChen

Hi Habal,

There are probably different ways to do this. One option is the remap tool in ArcGIS Pro, which can let you reclassify the pixel values from 15 to 5 by mapping a range of values to a new value.

https://pro.arcgis.com/en/pro-app/latest/help/analysis/raster-functions/remap-function.htm#:~:text=The%20Remap%20function%20changes%20or,values%20to%20output%20pixel%20values.

Hope this helps,

-Jay