We need the ability to not only detect text in images but also to remove it or at least get the coordinates in the OCR data so we can remove it ourselves.