adf2pdf 0.8.0

Automate the workflow around ADF scanning, OCR and PDF creation

adf2pdf - a tool that turns a batch of paper pages into a PDF with a text layer. By default, it detects empty pages (as they may easily occur during duplex scanning) and excludes them from the OCR and the resulting PDF.

For that, it uses [Sane’s][5] [scanimage][6] for the scanning, [Tesseract][4] for the [optical character recognition] (OCR), and the Python packages [img2pdf][9], [Pillow (PIL)][10] and [PyPDF2][11] for some image-processing tasks and PDF mangling.

Example:

$ adf2pdf.py contract-xyz.pdf

2017, Georg Sauthoff <[e-mail]>

## Features

• Automatic document feed (ADF) support

• Fast empty page detection

• Overlaying of scanning, image processing, OCR and PDF creation to minimize the total runtime

• Fast creation of small PDFs using the fine [img2pdf][9] package

• Only use of safe compression methods, i.e. no error-prone symbol segmentation style compression like [JBIG2][12] or JB2 that is used in [Xerox photocopiers][12] and the DjVu format.

## Hardware Requirements

A scanner with automatic document feed (ADF) that is supported by Sane. For example, the [Fujitsu ScanSnap S1500][1] works well. That model supports duplex scanning, which is quite convenient.

## Example continued

Running _adf2pdf_ for a 7 page example document takes 150 seconds on an i7-6600U (Intel Skylake, 4 cores) CPU (using the ADF of the Fujitsu ScanSnap S1500). With the defaults, _adf2pdf_ calls scanimage for duplex scanning into 600 dpi lineart (black and white) images. In this example, 6 pages are empty and thus automatically excluded, i.e. the resulting PDF then just contains 8 pages.

The resulting PDF contains a text layer from the OCR such that one can search and copy’n’paste some text. It is 1.1 MiB big, i.e. a page is stored in 132 KiB, on average.

## Software Requirements

The script assumes Tesseract version 4, by default. Version 3 can be used as well, but the [new neural network system in Tesseract 4][8] just performs magnitudes better than the old OCR model. As of mid 2018, there is no stable version 4, yet, but since the beta version is so much better at OCR I can’t recommend it enough over the stable version 3.

Tesseract 4 notes:

• [Build instructions][2] - warning: if you miss the autoconf-archive dependency you’ll get weird autoconf error messages

• [Data files][3] - you need the training data for your languages of choice and the OSD data

Python packages:

• [img2pdf][9] (not packaged for Fedora, yet) - version 0.2.4 works fine

• [Pillow (PIL)][10] (Fedora package: python3-pillow-devel)

• [PyPDF2][11] (Fedora package: python3-PyPDF2)

[1]: http://www.fujitsu.com/us/products/computing/peripheral/scanners/product/eol/s1500/ [2]: https://github.com/tesseract-ocr/tesseract/wiki/Compiling-–-GitInstallation [3]: https://github.com/tesseract-ocr/tesseract/wiki/Data-Files [4]: https://en.wikipedia.org/wiki/Tesseract_(software) [5]: https://en.wikipedia.org/wiki/Scanner_Access_Now_Easy [6]: http://www.sane-project.org/man/scanimage.1.html [7]: https://en.wikipedia.org/wiki/Optical_character_recognition [8]: https://github.com/tesseract-ocr/tesseract/wiki/NeuralNetsInTesseract4.00 [9]: https://pypi.org/project/img2pdf/ [10]: http://python-pillow.github.io/ [11]: https://github.com/mstamy2/PyPDF2 [12]: https://en.wikipedia.org/wiki/JBIG2