About two years ago, I did some preliminary footwork into the feasabilty of using some of the current technologies in face/optical/photo recognition and applying it to identifying chips, specifically to chart and categorize eBay chip sales. The idea was to develop a program that could read all the closed auctions and chart/categorize info based on the chip photo. At the time, I had been charting the sales of chips that sold for over $25 by hand but it some became overwhelming. I then had an accomplished software writer (a reative, who did at for "fun") create a program to sort the sales but it failed miserably in placing the chips into issues. It had to be done by hand. Next step was sorting based on the photo. We looked into it.

In short, there are many hurdles, some legal, and at the end of the day, developing the specific software to accomplishg the task is expensive with a very limited market and almost no chance of realistically recouping the investment needed to develop it. People I spoke with said it could be done, but not for under six figures.

Plan B was to get a university to develop it as an interesting project. I pitched it a few times, with no success. In some discussions, I seem to recollect the club trying (or considering) to pitch the idea to a university, but I don't know if it was ever pursued.

Compiling your database of photos for comparison is a gargantuan task too, and you need permission for any of them you don't own.

In short, the prize was not worth the contest. We gave up. Maybe the landscape has changed since....