You are reading a single comment by @allObjects and its replies.
Click here to read the full conversation.
Espruino is a JavaScript interpreter for low-power Microcontrollers. This site is both a support community for Espruino and a place to share what you are working on.
@allObjects
I m not an expert in the matter, nor do I know the algorithms.
But the triple take and some calculation makes sense to me... for different reasons: first, enrollment has to show 3 times 'the same' - the same with 'tolerable variation'. Second, when validating, the template with variation is used to detect 'identity'. The triple take allows to calculate a mean and some deviation to be stored in a 'verification matrix' which then is used on identity check.
You can give it a shot by messing with the enrollment... I'm pretty sure, if you do not use the same finger for all tree takes, enrollment will show error. But it would be interesting to see how much 'has to be the same' or 'present' over all 3 takes to make it still a successful enrollment, and at the reverse: how much can be different - not just 'absent'.
The same is for checking identity: how much can be absent, how much can be different...
When this technology was introduced at the US borders, we - my wife and I - entered in some northern station middle in the winter. For me the readers worked perfect... for my wife not at all... with winter dry and 'cold' hands/fingers, the reader 'failed' completely... not even warming up with under warm water did help. In the end they let us pass without being able identify by finger print reader... It is know that some refugees burn / mutilate their skin to make finger print identification fail... for different reasons...