Thanks Vasily and all your work on this.
I actually had the raw image dumped in fpi_assemble_lines instead and there
is no more distortion. I tried to debug the exact issue but the code kind
of makes my head spin. I think maybe not all images should be processed
through this function, instead only "fast swipe" images that are shorter
than a certain tolerance should be. Just a thought. If I come up with
something, I'll send it over.
On Fri, Dec 16, 2016 at 7:12 PM, Vasily Khoruzhick
wrote:
> Hi Carson,
>
> All related code is in libfprint/assembling.c, you're probably
> interested in fpi_assemble_lines().
>
> Regards,
> Vasily
>
> On Fri, Dec 16, 2016 at 1:17 PM, Carson Reinke wrote:
> > I seem to be having the same issue as reported by other awhile ago here:
> > https://github.com/ars3niy/fprint_vfs5011/issues/9. The head from
> master
> > seems to work better, but still quite a bit of distortion.
> >
> > I believe this from trying to normalize the image from fast swipes. Can
> > anyone point me in a direction on where this is done?
> >
> > ___
> > fprint mailing list
> > fprint@lists.freedesktop.org
> > https://lists.freedesktop.org/mailman/listinfo/fprint
> >
>
___
fprint mailing list
fprint@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/fprint
