Project Transcribe v 1.0
Project Transcribe v 1.0 was originally titled 'The Ugaritic Transliterator' and was a project undertaken by both Petros Koutoupis and Garrett Hirschey while attending DeVry University. The original goal of this project was to input an image of an ancient script from an external source (in an image readable format such as *.jpg, *.bmp, *.tif, etc.) and use Digital Signal Processing (DSP) to recognize the script and compare it to a library of ancient characters (relating to that script). This in turn would be transliterated to its true phonetic value in Roman characters, using Unicode-based transliteration syntax.
Originally, the ancient language tested and used was Ugaritic and due to the many difficulties of recognizing hand written text/characters, it would have been too difficult to work with tolerances of image recognition (14 weeks was given to complete this); so it was chosen to work in an ideal environment with already ideal characters, setting a basis to work off of in the near future. The project was completed successfully.
Aside from being able to transliterate ancient script from either scanned or imported papyri, cuneiform tablets, etc., we were aiming to do much more. This idea came to mind when Petros Koutoupis came to the realization that there are millions of ancient scripts that have gone un-deciphered, hidden in the basements of museums and galleries, to even, unfortunately, private collectors; and scripts that need to be re-deciphered for the purpose of re-understanding ancient cultures and beliefs from a new 21st century perspective; and not the objective 18th to the early 20th views based on Biblical sources. The reader has to understand that in the early stages of archaeology, the only history we knew was what was written in the Biblical canon and also from various classical (ancient) Greek and Latin writers; this in turn may have led to incorrect and biased assumptions. The technology is here and capable. Why are we not taking advantage of this?
Here is a rough draft of the GUI outline(s):
The exact meaning to transliterate out of the English dictionary is: To represent (letters or words) in the corresponding characters of another alphabet.
When something is being transliterated into the Roman script, the individual is reading the word in the original language; but using characters most familiar to them. It is by this process that almost all scholars go by in order to properly document and translate ancient writings. Once a string of characters have been identified, a dictionary would be thrown onto the scene in order to properly translate what has already been transliterated. Please see the example below:
Here is an example of the original Ugaritic alphabet we used with its corresponding Hebrew and Romanized phonetic sounds:
As of 8-Oct-2005, this is where the project stands:
1. This is the Ugaritic Transliterator v.1.1 GUI (Graphical User Interface) loaded up for transliteration. |
2. To load the image, you have to click File and then Load Source.
|
3. Select the file to be tranliterated. |
4. It will then be displayed on the left window.
|
5. To execute the transliteration process, go to the Transliterate option on the top menu bar and select Execute. |
6. All recognized characters will be transliterated using Unicode syntax and displayed on the right window.
|
Note that this was the demo submitted by both Petros Koutoupis and Garrett Hirschey while attending DeVry University.
No work has continued on this project since.
With the recent shift toward mobile computing, it has become much easier for developers to build applications to run on those same mobile platforms. A good example would be today's smartphone (2010). With Operating Systems such as Google's Android, the possibilities are limitless and it would be very simple to build a similar framework to run on a phone which will utilize the phone's built in camera for real-time transliteration and translation. With the appropriate funding this can be accomplished in a short time frame and for those interested parties that wish to learn more, please contact me by e-mail. It is worth noting that the possibilities for this application are endless. While the demo shown above was intended for dead languages, this future application suite would support most of todays commonly used and studied languages. If one were to take a trip to Greece, Japan, Israel, Egypt and more, they would be able to pull out their smartphone, capture the text into the camera's viewfinder and wait for an identification, transliteration to even translation of the text in question. This could be for both commercial and military use.
| 
