I'm trying to find a way to screen scrape the letters and numbers (mainly numbers) from the attached picture.
In previous attempts, I've used pyocr and many other variations.
My question is, has any body found a way to scrape off numbers? Or how to train the pyocr algorithm to use custom data?
Thanks in advance!
The folks at PyImageSearch have a TON of info about processing images in Python with OpenCV.
They even have a free blog post about using Tesseract OCR. Though Tesseract can be a bit fussy about fonts, the good news is it looks like your text in the image should always be the same font, and perfectly aligned horizontally and vertically.
(disclaimer: I'm a student of theirs; but I don't work for them) https://www.pyimagesearch.com/2018/09/17/opencv-ocr-and-text-recognition-with-tesseract/
