from Hacker News

Ask HN: Google Vision vs. AWS Textract vs. Tesseract?

by takkatakka on 12/22/23, 5:20 PM with 0 comments

I've been playing around with Tesseract to extract text from pdf files and have found it a little disappointing, even after doing some pre-processing to the image (300dpi, binarization, de-skewing).

Does anyone have good experiences with OCR tech for extracting text from PDFs? Especially scenarios where you can't control / improve the initial image.