|
HI all, I am trying a classification in SEPAL using a table from Earth engine as training data Link: to GEE asset: https://code.earthengine.google.com/?asset=users/matteojriva/timor/Atauro_collected In earth engine the collection displays well, However in SEPAL I get the following error "Earth Engine: Image.classify: No data was found in classifier training input." Is there any requirement for the training class or dataset? should they be string or integer or of any particular form? EDIT: solved the problem. The problem was that the radar image was empty (cloud filtered) thus there was no value for the training point. Incidentally I also realised my SEPAL account was not connected to my GEE account, limiting the possibility to export image from SEPAL to GEE Thanks |
|
The column for the training data should be in integer format |
|
SO the problem must not be that as it is already in integer format. Any other idea on what it could go wrong? Should I be able to see the point in the SEPAL preview? |
|
Dear ciskoh The problem was that the radar mosaic you used was empty, we solved by including both descending and ascending orbits to your recipes. Also, make sure your SEPAL is connected to your GEE account if you want to export recipes and classifications as GEE assets |