Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Small sensors cannot use minutiae #4

Open
iafilatov opened this issue Mar 10, 2018 · 3 comments
Open

Small sensors cannot use minutiae #4

iafilatov opened this issue Mar 10, 2018 · 3 comments

Comments

@iafilatov
Copy link

Now it becomes very difficult for the algorithm to find MINUTIAE.

Because often there's nothing to find on such a small area. libfprint wants at least 10 to enroll but it's a very liberal criteria and assumes you always carefully place your finger exactly the same way on the scanner. Verification will also want 10+ but then it's really hard to get a good match score. From my experience, if you can't reliably get 30+ minutiae, using the scanner is not a pleasant experience. Pure minutiae based algo just isn't made for modern-day small scanners and libfprint really needs to have another one for them.

The number of pixels per inch is hard coded: DEFAULT_PPI = 300

This is because actually tweaking it doesn't seem to change anything. If you follow the parameter through nbis code, it seems to only affect how minutiae quality is estimated.

.flags = 0, // What is this?

Not used for anything really.

// Maybe we should move open to activate?

I would. Most drivers use dev_open for configuration and dev_activate for initialization that requires talking to the devices. It's not a rule though.

As for why enroll fails, I'm not sure. I recommend heavy use of fp_dbg (see autogen.sh). Please post examples/enroll with debug output.

@hmaarrfk
Copy link
Owner

Thanks Igor,

The comment about minutiae was along the lines of what I was thinking. I think developing is a different endeavour compared to obtaining images from the vfs7552.

Regarding open/activate: I assume the deamon would have to re-initiate the sensor after waking up from sleep. Therefore, I'll move the communication with the usb device to "activate" in case the daemon actually calls activate when the laptop wakes from sleep.

@iafilatov
Copy link
Author

I don't know how wake up is handled in general. I'm using vfs5011 and elan and suspending constantly. The former does stuff in open, the latter in activate. No prob on wake up. I think either way is fine.

@hmaarrfk hmaarrfk changed the title Some comments Small sensors cannot use minutiae Mar 10, 2018
@hmaarrfk
Copy link
Owner

Alright good to know. I think this is going to be an edge case, where the user closes his laptop lid during an enrolling process.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants