Champollion 2.0 was born from a need I had: I began my own genealogy in 1990. There was no electronic document at that time, but into the 2000's, the need for specific software that would help handle historical documents started to grow. During the year 2012, I started programming one small utility. I then added features as I encountered new needs. In 2014 my program became a full application, and I decided to work harder and try and turn it into a fully grown commercial application. I started my company in late 2014 and started to sell the software in February 2015.
What it does
Champollion 2.0 is a powerful program for windows that provides a wide set of tools for everyone who works with historical documents:
Specific image processing
- Simplified and automatic settings
- Wide set of contrasts settings
- Transparency handling
- Crop, skew, and correct parallax issues
- Quickly and easily number the lines of your manuscript
- Non-destructive annotation straight on your manuscript
- Automatic picture backup: safely try several settings
- Image comparison between the original and processed images
Specific word processing:
- Make text and image interact with each other
- Handles multi picture documents: Include "next folio/next vue" marker in your text
- Searchable, multi criteria database for your documents: combine historical date, place, full text search to retrieve your documents in a wink
- Batch image processing
- Batch file handling (smart renaming, moving, copying, deleting)
- Batch PDF images importation (handles all embedded images formats)
Multi purpose help:
- Hovering icons displays a short text
- Interactive tutorials
- User's guide
- Contextual help
- Responsive technical support included
How I built it
I used my engineer’s expertise in programming and image processing, and I programmed it in standard languages for Windows: Visual Basic, first, then I had to move to C# and even C++ and C for more efficient image processing libraries.
Challenges I ran into
One of the hardest challenge I had to deal with was to achieve a smart ink detection on the manuscripts. I'm using a self-made algorithm that uses grey-level gradient analysis and reconstruction scheme. And it does work just fine, whatever the type of manuscript and their quality. The algorithm allows the user to get rid of any bleed-through on any manuscript. Another one I'm proud of was to connect my application to the internet: automatic updates, active embedded technical support... it was worth the work.
Accomplishments that I'm proud of
Champollion 2.0 is starting to be famous in France now. Several newspapers and magazines have already published positive reviews.
What I learned
This project lead me into C# programming. A great language I wasn’t used to, which is somehow combining the advantages of C/C++ with the ease of BASIC. I also learned how to pay attention to suggestions from users. It lead me to include technical support directly inside the software.
What's next for Champollion 2.0
Champollion 2.0 will bring much more active help for palaeographers. Next step - available during 2017 - will be to let the users build and share their own palaeographic references. First tests are pretty much promising...