I've noticed that when importing a recipe from a website that the system auto scans the text and pre-populates the cuisine region automatically but it does not do this with scanned images or even when creating a new recipe from the text you type in. It would be nice if that feature was present on all input methods.