robinxl / pdftoimage

Yet another pdf to image

pdftoimage

A light weight python(3.6+) module for converting pdf to image. Wrapped with GhostScript

dependant

Install ghostscript 9.26

Mac

brew install ghostscript

CentOS/Fedora

sudo dnf install ghostscript

Ubuntu/Debian

apt install ghostscript

Verify ghostscript

gs --version
# output: 9.26

How to use

Use as command line

python pdftoimage.py test.pdf test

# output:
Output directory not exists, create ...
Image created at: test/test_page1.jpg
Image created at: test/test_page2.jpg

Bulk convert all PDF to image in the folder (includes all subfolders)

python bulkconvert.py test

# output:
Found total 3 PDFs
Output directory not defined, use the same directory with input PDF ...
Output directory not exists, create ...
Image created at: ./test_PDF_TO_IMAGES/20200109174921-0001.jpg
Error: Ignoring invalid annotation, output may be incorrect.
Image created at: ./test_PDF_TO_IMAGES/8-15-15 A320J158001-RW_page1.jpg
Image created at: ./test_PDF_TO_IMAGES/8-15-15 A320J158001-RW_page2.jpg
...
Image created at: ./test_PDF_TO_IMAGES/0012063430.jpg
Job finished! Total time takes: 7.694029092788696

Working with PIL or OpenCV

OpenCV / PIL

import pdftoimage

myPDF = 'example.pdf' # or byte type pdf

# OpenCV
cv_images = pdftoimage.get_cvimages(myPDF)

# or PIL
pil_images = pdftoimage.get_pilimages(myPDF)

Future work

• Image format flag
• DPI flag
• OpenCV flag
• more option on output image control (grayscale, denoise)