Illegible words on church headstones could be readable once again, thanks to a scan technology developed at Carnegie Mellon University. The new technology creates high-resolution, 3D scans of tombstones to reveal the carved patterns in the stone. A computer matches the patterns to a database of signature carvings, which reveals the words.
In recent weeks, a research team has been testing the new technique at a 200-year-old cemetery near the university. The pastor at Old St Luke's Church hopes to identify all the names on the cemetery's tombstones.
The technology has been built on top of existing 3D reconstruction techniques. But the team at Carnegie Mellon is focusing on surface signatures. "We have designed special filters of 3D data that can detect curvature or linear features on a surface," explains Dr. Yang Cai, director of the Ambient Intelligence Lab at Carnegie Mellon Cylab.
You can read more at http://news.bbc.co.uk/2/hi/technology/7024672.stm. I found the "before-and-after" pictures on that web site to be very impressive.
I know a few gravestones that would be handy for. I find most interesting the comment: "The technique could one day also be used by doctors to examine a patient's tongue for signs of illness." Doctors in China have already been doing that for 5,000 years.
http://www3.telus.net/Jgen/jlog.html
Posted by: JLB | October 05, 2007 at 03:10 AM
I could use that technology NOW!
Posted by: Margaret | October 06, 2007 at 09:46 AM
Margaret, I'm with you on that! I have a cemetery right down the street that I have worked hours and hours on trying to identify text on some of the stones. I've taken pictures at many different times of days, tried all kinds of filters, and information is still lost.
Posted by: Cheryl | October 06, 2007 at 11:58 AM
So what is the equipment involved - the article doesn't say. Do they use handheld scanners that are outfitted with the program, or is the program located on computers in their lab, and they use two scanners, one high-resolution, and one low-resolution?
I would guess this would be beyond the means of most genealogists, especially if the software remains proprietary and they can sell it to doctors, etc., for big bucks.
Posted by: Susan Daily | October 06, 2007 at 03:03 PM
This reminds me that in 1995 I knew in time we would be able to read unreadable stones. I was visiting in Wheatland, WYO at my Uncle's and we had an old unreadable glass picture in a case. We were sure whom the picture was of because we had the readable pictures of the other siblings from the Civil War. We originally did not take it down to make copies at the Copy store my Uncle used. The owner sent my Uncle back to the house to get the glass picture. Low and behold when we put that black glass picture on his laser
copier,printer at the store out came the picture of my ancestors brother whom we were sure it was of. It was explained that the laser bounced off the background thereby recreating the picture that on the outside had turned black.
Later that month I was in Afton, Wyo visiting a cousin and the light on the mountainside made some stones more visible some less. You could see when the light hit the stone just right that the impression (for lack of better description)was still in the stone. It was just to deep to get a picture( technically to shallow on the stone)to read. I had already shared with her the picture of our GGUncle from Wheatland collection. We felt in time someone would be able to do the same to tombstones that others were doing to glass pictures. It was a matter of how to do it without stone damage.
It makes sense that if you hammer a name or date into stone that stone was altered in compressions plus some stone was removed. That impression into the stone should move on a laser system the same as on the picture.
SusiCP
Posted by: Susan Pentico | October 06, 2007 at 04:29 PM
A word of warning. NEVER take a black picture apart it destroys the data you need to make the picture. SusiCP
Posted by: Susan Pentico | October 06, 2007 at 04:31 PM
An article regarding this project, with pictures during a "session," appeared in the Pittsburgh Tribune Review today. http://www.pittsburghlive.com/x/pittsburghtrib/news/cityregion/s_531518.html
Posted by: Tom | October 08, 2007 at 10:15 PM