Convert PDF files to PNG images

This robot converts PDF files to PNG images using Python. This is useful when you want to use OCR (Optical Character Recognition) and image recognition services to extract data from your documents.

There are two example PDF files:

• A single-page example-invoice.pdf
• A multipage example-multipage.pdf

The pdf2image library generates one image per PDF document page:

example-invoice.pdf-0.png
example-multipage.pdf-0.png
example-multipage.pdf-1.png
example-multipage.pdf-2.png
example-multipage.pdf-3.png
example-multipage.pdf-4.png
example-multipage.pdf-5.png
example-multipage.pdf-6.png
example-multipage.pdf-7.png
example-multipage.pdf-8.png

Dependencies

conda.yaml:

channels:
  - conda-forge

dependencies:
  - python=3.7.5
  - poppler=21.03.0
  - pdf2image=1.15.1

The robot

task.py:

from pdf2image import convert_from_path


def convert_pdf_to_images(pdf_path):
    images = convert_from_path(pdf_path)
    for index, image in enumerate(images):
        image.save(f'output/{pdf_path}-{index}.png')


def task():
    convert_pdf_to_images('example-invoice.pdf')
    convert_pdf_to_images('example-multipage.pdf')


if __name__ == "__main__":
    task()