On Mon, Sep 29, 2003 at 11:10:42PM -0400, David Hobby wrote: > I wasn't that worried about resolution, just signal to noise > ratio.
They are related. Good luck resolving fine details with a low S/N. > The other issue that comes up is having the system target > the eyes of a moving person. Moving the system to track the > person seems difficult. So that leaves the other option, which is > to constantly record the image from anywhere the eyes might appear, > at a resolution fine enough to identify the eyes. NOW you're talking about resolution! > I don't believe the optics are that big an issue, since they > could be designed to only work well at one wavelength. A lot of > the price of good lenses is to make them achromatic, focusing many > different wavelengths to almost the same location. Is an iris pattern still as unique if you only look at one wavelength? But anyway, I guess the answer may be to use the lenses that have already been designed and produced for consumer megapixel cameras, those are apparently of reasonable quality. Anytime you can use parts from a high-volume consumer part, you can save a lot of money. > "Seeing" might be an issue. I doubt that you could get better > than 1 arc second resolution in a public place, the air would be too > turbulent. But taking my 1 mm resolution criterion, we get that > the camera could be as far as 200 meters away, since 1 mm spans around > 1 arc second at that distance. That should give some room to place > cameras in... And I doubt you need as much as 200m. In most public places, you could hide them closer than that. -- Erik Reuter http://www.erikreuter.net/ _______________________________________________ http://www.mccmedia.com/mailman/listinfo/brin-l
