Vishal Rao wrote: > OK, I increased the MAX_FRAMES from 150 to 350 to avoid "swiping too > slow?" error, then I press firmly and swipe really slow. > > I now get better scans (attached) and the minutae count is 10 to 20 > but the bozorth score is anywhere from 3 to 12 which I guess is not > enough for a match value of 15?
Indeed. The scans look really weird and quite different from Anthony's (which are much clearer). I am not surprised you are not getting decent image processing performance. We have a similar bug for aes2501 where it works for seemingly everyone except one person: http://www.reactivated.net/fprint/bugs/task/2 The images he gets are very different. The windows AES4000 driver does dynamic recalibration of gain depending on the pressure the user is applying to the device. I am fairly sure that the same is true for AES2501 and AES1610. It sounds like this dynamic recalibration is also useful to handle devices which clearly differ in some way. None of the libfprint authentec drivers do any dynamic recalibration at the moment, but when we've implemented that, it should make problems like yours go away. Please file a new bug for your issue at http://www.reactivated.net/fprint/bugs/ Thanks for your feedback. Daniel _______________________________________________ fprint mailing list [email protected] http://lists.reactivated.net/mailman/listinfo/fprint
