Thanks for reading! If the text in a PDF file is OCRed, then tabula should work. If not, you need to OCR the text first. In that case, you can try with pytesseract.