I am trying to pre-process an image for OCR. How do I fill in the contours of the text to make it more readable for the algorithmenter image description here