![NXP Semiconductors SLN-VIZNAS-IOT User Manual Download Page 34](http://html.mh-extra.com/html/nxp-semiconductors/sln-viznas-iot/sln-viznas-iot_user-manual_1721903034.webp)
34
SLN-VIZNAS-IOT User Guide, Rev. 1.1, 10/2020
NXP Semiconductors
Figure 45: On-screen Info for Registering Face
NOTE: Without any registered users, the on-screen debug info will not update until a registration
has been triggered. This is because the inference engine does attempt to recognize any faces if
they are none currently registered.
In the above example, registration is taking place, but according to the on-screen
debugging info, although the face has passed the blurriness check (hence
“blur: 0”
), the
“front face check” has passed (hence
“front: 1”
), and the RGB liveness check has
passed (hence
“rgbLive: 1”
), the IR liveness check has not passed (hence
“irLive: 0”
).
This would suggest that there may be an issue with the IR camera
– likely a lighting issue.
Figure 46: On-screen Info for Recognized Face
In the above example, the on-screen info reflects information about the face being
recognized. As shown in the screenshot, the face has been recognized with 95% similarity
to the matching face in the internal database (hence
“sim: 95”
) and has passed both the
RGB and IR liveness checks (hence
“rgbLive: 1”
and
“irLive: 1”
).
NOTE
: The “front face check” is only performed by the inference engine for faces being registered
to ensure that no face gets registered that it is improperly angled towards the camera.