wxCover wrote: > I've installed both from git and tested all that stuff. > There is a difference between the scores. I've changed bz3_threshold to > 15. With this setting, it works better: I get many more matches. I > haven't seen any wrong positives with that. But there are some false > negatives. Maybe we could lower bz3_threshold a bit more? When I read > another finger than the one enrolled, I never get scores higher than 7. > The maximum score I got was 32.
OK. The images look nice, and the minutiae detection has worked OK too. It must be that the reduced size of this sensor simply sees less of the finger. Everything looks great but the image processing code is not performing quite as well as we would hope. I guess this is because it is tuned for larger numbers of minutiae. AES4000 has a similar problem -- it works, but not anything like as well as aes2501/uru4000. If you have time and interest to look into this, please dive right in :) I am planning to look at it sometime but have other priorities first. > I've tried to change the enlarge factor but it didn't help (in fact, it > may be worst=scores lower). Maybe because the binarization gives pretty > good results (see the attached files)? Indeed, that looks fine, so enlargement won't help. I'll post some pictures of what enhancement+binarization did to AES4000 sometime... > In fact a big issue is to have a correct enrollment. I think in the > future we'll need to read several times the same finger and see if > there's no problem with it. Anyway, you may already know that! Yes, I'm planning for a 3-swipe image enrollment system where we pick the best image (ranked first by a quality algorithm, and then by minutiae count). I am going to finish off a few tasks and then do some new releases of all 3 components. Thanks for your work! Daniel _______________________________________________ fprint mailing list [email protected] http://lists.reactivated.net/mailman/listinfo/fprint
