avi

In my opinion, DAZ, Stephan is well-intentioned, here, I certainly do not imagine that he is out to fleece anyone.

However, this approach is completely wrong.

One does not inquire about the CV of the three students.

One does not require any students.

In fact, the scrutiny of 10,000 digital images, seeking a pattern match with another digitial image, is not a task for any human being, that's why God invented computers.

avi

stephan huller

Judge

The reason the number is 500 Euros off is because I am going to contribute that much

Avi

The idea of getting a computer to make the comparison is intriguing but I don't believe it's possible. Who would I contact to undertake such an activity?

Toto

Computerized handwriting analysis is a relatively new field. It might take much longer to develop the algorithms for 18th century Greek handwriting and validate them than to have 3 graduate students do the scan manually. But it might make the whole project more interesting to an investigator.

But if you are interested in that approach, that link references There is a paper by Srihari here from 2003.

Crime labs use FORENSIC INFORMATION SYSTEM FOR HANDWRITING (FISH)

stephan huller

Thanks avi and Toto

I have sent an email to Srihari and I will let you know what he says.

avi

There is nothing magical here, Stephan.

It is very easy. It is a digital image, not a sample of "Greek handwriting". It serves as the template.

The ten thousand other digital images are compared, pixel by pixel with the template.

Some few dozen images will demonstate overlap.

The task then is to give scrutiny to those 30 or 40 images which are candidates for a match.

The template, here, is indeed Greek handwriting, not a digitized musical score, or an image of a transmission component from an automobile engine, or a map of bosnia.

But the principle is the same. Template image either demonstrates some features similar to the item under investigation, or it does not.

This method, digitial image analysis, is not specific to ancient greek manuscripts. It works for any kind of digital images.

The field of optical pattern match is more than thirty years old. There are many, many investigations published, in several different journals. There are also books on the subject. Remember, this has nothing to do with any religious texts. The method of investigation doesn't differentiate human handwriting, from lunar landscapes. There is a template, which has a pattern of pixel illumination that either conforms, to some extent, with each succeeding sample, or it does not. One is left with a handful of candidates to examine further....

I can identify, by private message, a member of the forum, who is very knowledgeable about this process, if you wish to learn more.

avi

stephan huller

Sure please contact me with that information. Thanks avi

Vorkosigan

Steven, can I do it by Paypal?

stephan huller

Vorkosigan

Sure. I haven't arranged for the payment yet but when the time comes I will make sure to arrange for that. I am just looking for a verbal commitment and then I will probably get someone like the Society of Biblical Literature or Biblical Archaeology Review to handle the money collection. Can I put you down as a committed donor?

dog-on

Stephan, I'll donate.

stephan huller

thank you so much dog-on. Really appreciate it.