Data Science Stack Exchange is a question and answer site for Data science professionals, Machine Learning specialists, and those interested in learning more about the field. It only takes a minute to sign up.
Sign up to join this community
Problem Statement: Identify the sub-region in an image with hand-written text/scribbling
(Attaching an example for reference. Masked some text for privacy)
Once I get the pixel values, I want to replace them with white pixels, so I remove the scribblings completely from the image.
(PS: I don't have a lot of training images to train a CNN. So, asking here in case the problem can be solved via a classical CV approach)
You could try to run some OCR like tesseract. It'll give you locations of detected text. It'll probably fail for handwritten and skewed text. You'll get your mask, though this is a bit hacky as some false negatives may appear.
