PDFOCR

State-of-the-art OCR processor using the open-source Tesseract^[1] OCR engine. Generates searchable mixed-mode (dual-layer) PDFs from image-mode (single-layer) PDF input files.

This is a test pilot. The accuracy and availability cannot be guaranteed.
The test pilot does not operate in a production environment. In case of problems, contact the developers, not Helpdesk.
The OCR processor can convert huge PDF files with hundreds of pages.
Processing speed is about 3..10 pages/minute for the current implementation.
Jobs are queued: one job, i.e. one user is served per time, then the next.
You receive an automated notification e-mail including a download link to your processed PDF file when the OCR process has finished.
The first version ran on 13.06.2014.
Known issues which are already on the to-do list and which we want to fix later:

Processing failures like empty images or empty ocr result pages are currently not detected and not signalled to the user.
Protected files (example: "c't" archive files from the HEISE Verlag) cannot be processed currently, because pdftk refuses to split the pdf into single pages unless a password can be provided.
Currently, you can only set up one ocr language per upload.
Processed files are not automatically deleted. Please use the button to delete the files you have downloaded.
Bug: when you delete a file which is in progress, this process is not correctly terminated: user has to wait, until the file is processed (most likely this results in an empty file) and then to restart by resubmitting the file.
Nerds and statistic freaks might wish to have a usage log and statistics, and an indicator of how much energy is wasted by ocr-ing files (carbon footprint). Nice wish. Try harder.
The software is not yet ready for multi-core/multi-user processing.

image.pdf => mixed-mode.ocr.pdf

In addition to this online ocr conversion tool,
you can download a Java tool
for offline highlighting terms in your downloaded ocr-ed PDF files.
Simply download the tool to your desktop and unzip it once.
Start pdfhl.cmd (in folder pdfhl after unzipping).
Select the language of your document before you upload it:
A multiple selection is currently not possible, but planned.

image.pdf => mixed-mode.ocr.pdf In addition to this online ocr conversion tool, you can download a Java tool for offline highlighting terms in your downloaded ocr-ed PDF files. Simply download the tool to your desktop and unzip it once. Start pdfhl.cmd (in folder pdfhl after unzipping).	Select the language of your document before you upload it: A multiple selection is currently not possible, but planned.