Skip to content

Latest commit

 

History

History
101 lines (70 loc) · 4.33 KB

README.md

File metadata and controls

101 lines (70 loc) · 4.33 KB
Raw

fingerprint-reader-daemon

A daemon service allowing web apps to read data from biometric devices, currently working with fingerprint readers, tested with Digital Persona u.are.u 4500 and the Neurotechnology Standard SDK.

Disclaimer

THIS IS NOT READY FOR PRODUCTION, use it ar your own risk.

This work was done as a PoC, being a piece for a bigger project, there are several things that aren't implemented on the security side, like checking which websites are allowed to read the fingerprints, dealing with several devices connected, etc.

But, this can certainly help you to get started.

NOTE: Avoid using Windows VMs, we spent lots of time trying that but back then, there was a known bug on the SDK that prevented us from doing that, the workaround is to setup a license server which is tricky.

Run

While the project should work on linux (tried in ubuntu 18.04), we only have instructions for Windows.

You will require:

  • Java 8 (it may work with newer versions but we haven't tested them).
  • Gradle

Run for windows

First, you need to install the SDK from Neurotechnology Standard SDK, we tried Neurotec_Biometric_11_1_SDK_2019-07-07 but should likely work with newer versions and we suggest you to use the latest one.

Everything must be run the as an administrator.

Save in $SDKDIR your Neurotec_Biometric_11_1_SDK path, for example (replace your SDK path):

  • For powershell, run Set-Variable -name "SDKDIR" -Value "C:\Users\jonsa\Desktop\neurotec\Neurotec_Biometric_11_1_SDK"
  • For cmd, run set $SDKDIR="C:\Users\youruser\Desktop\neurotec\Neurotec_Biometric_11_1_SDK"

Run the SDK activation service

It is necessary to run the activation licenses in the background, once the SDK is downloaded and uncompressed:

cd $SDKDIR/Bin/Win64_x64/Activation
./pg.exe -install

You can stop the service with ./pg.exe -uninstall

Run the daemon

Be sure to keep your fingerprint reader connected, and make sure that the necessary drivers are installed, you can try the SDK examples to make sure that your computer can read the fingerprints.

We assume you already cloned our project, then:

  • Create the directories Bin/Java
  • Copy the necessary libraries from the official SDK (on $SDKDIR/Bin/Java to Bin/Java)
neurotec-biometrics-client.jar
neurotec-licensing.jar
neurotec-biometrics.jar
neurotec-media.jar
neurotec-core.jar
neurotec-media-processing.jar
neurotec-devices.jar

After that, gradle run should be enough.

Troubleshooting

If you get a License error problem you must copy the dependencies from the SDK, for example:

cp $SDKDIR/Bin/Win64_x64/*.dll enroll-finger-from-scanner/build/classes/java/main
cp $SDKDIR/Bin/Win64_x64/FScanners/NdmDigitalPersonaUareU/*.dll enroll-finger-from-scanner/build/classes/java/main
cp -r $SDKDIR/Bin/Data enroll-finger-from-scanner/build/classes/java/main

When you see the fingerprint reader connected but it's not detected by the app, it's likely a driver problem, be sure to install the official custom drivers from Neurotechnology, running one of the official examples is the best way to make sure your environment is setup properly, like Neurotec_Biometric_11_1_SDK/Bin/Java simple-faces-sample.bat, if this example can't read a fingerprint, this app won't be able either.

Get a fingerprint

You can use any http client (like postman, httpie, or even your webapp) to execute the following request POST localhost:1212/fingerprints, which will read a fingerprint from the connected device.

Which initializes the fingerprint reader and waits until you capture a fingerprint, once done, it returns the base64-enconded image and it's template, like:

{
    "data": {
        "image": "aabbcc......",
        "template": "aabbccdd...."
    }
}

When there is an error, you should get a response like:

{
    "error": "You are already trying to read a fingerprint, try after completing that one"
}

The handled errors are:

  • Trying to read a fingerprint when there aren't devices available.
  • Trying to read a fingerprint when you are already reading another fingerprint.
  • Reading a fingerprint timed out, you took too long.