Is there a simple and fast way anyone can compare the two images with the new computer technology, or does it have to be done by professionals?
Helen, the more expertise you have in photography and in imaging, the better and more accurate the results from image enhancement. Also, when enhanced images are used in court, there are strict guidelines as to what is admissable, as it is very easy to alter an image into inaccuracy.
NASA's Retinex technology is available under the commercial name PhotoFlair. There are stand-alone and PhotoShop plug-in versions.
http://www.truview.comHowever, you would need a high-resolution scan of the original, and it is quite possible that the original does not contain enough data to work with. The negative would contain the most data, since the image could have been improperly focussed or over-esposed in the printing process. It's been maybe 15 years since my days in the darkroom, but I seem to remember negatives being sandwiched between glass, which had to be kept scrupulously clean or they would also blur the image. But since the NOVA documentary didn't even use the original photograph, it's doubtful that they had access to the negative, if it still exists.
Every time a photograph is copied, there is "generational loss" or "generational degradation," due to the random pattern of silver halide crystals that make up photographic film. In one 1950's military study of generational loss, it was determined that by the fourth generation (a copy of a copy of a copy of the original), photos taken by US spy planes were no longer useable; that is, the images had degraded to such a point that they no longer showed tanks, artillery etc.
If you're working from images scanned in from books, you're working with less information than even a fourth-generation copy. In books, images which were originally formed by crystals of silver halide in a random pattern are created by a regular pattern of dots. In an expensive art publication, these dots are relatively close. In books where the text matters more than the illustrations, the dots will be spaced farther apart.
I don't know that much about facial recognition systems, as it is a subset or superset of pattern recognition, and at present, only tangentially related to what I do for a living. All I've read is theory, and I have no knowledge of specific applications.
Here is a list of face recognition vendors:
http://www.biometritech.com/features/roundup051502.htmhttp://www.frvt.org/FRVT2002/Participants.aspMIT has been a pioneer in this area, and might be interested.
http://vismod.media.mit.edu/vismod/demos/facerec/index.htmlthe computer-generated match between photos of AA and FS
Elisabeth, this is not a computer-generated match. What the computer generated was a transition commonly used in movies and television. A dissolve: fade-out, fade-in.
A "computer-generated match" would be face recognition technology.
