Facial recognition systems can vary in how they function. Here is a general outline:
Face Detection
Facial recognition systems can identify a face, either alone or in a crowd. Current facial recognition systems can also scan faces whether they are looking straight at the camera or not.
Facial Analysis
When an image of a face is captured and analyzed, facial recognition solutions can search databases and identify matches. Facial recognition software can search a database looking for matches based on the distance between eyes or cheekbones – up to 80 nodal facial points.
Converting an Image to Data
A faceprint is created from the transformation of a face through mathematical formulas. Similar to the unique structure of fingerprints, each person has their own specific facial features that allow them to be recognized.
Finding a Match
A faceprint can then be compared against a database of facial photos. This process can happen in real-time or asynchronously. Using mathematical algorithms that take unique facial features into account, like bone structure and skin tone, a match can be found.
