It seems that the computer vision technology behind facial recognition is good for more than just recognizing faces. This time around, it’s being implemented to digitally reconstruct ancient artifacts.
The artifact in question is the Cairo Genizah, a collection of thousands upon thousands of texts written in Aramaic, using the Hebrew alphabet. Jewish tradition dictates that because Hebrew is the language of God, the writings cannot be destroyed, so the documents – which date back to around 870 AD – have been hidden for hundreds of years.
Image courtesy of the American Friends of Tel Aviv University
As scholars and travelers began to uncover bits and pieces, the writings – which consist of religious texts as well as practical record-keeping – have traversed the globe, and now pieces of the Cairo Genizah are scattered across the world’s libraries.
Rather than risk transporting the fragile documents, computer vision technology will make it so that they can be pieced, or “glued,” together digitally, by analyzing the handwriting and contents of the pages.
The Friedberg Genizah Project, a non-profit organization, has been digitizing all the various documents. Meanwhile, the scientists at Tel Aviv University are compiling them in order. They will announce what they’ve done in Barcelona, Spain, next month, during the 2011 International Conference on Computer Vision.
ComputerVisionSoftware 