What does it mean, Face Pattern Recognition?
Face recognition is a feature of the SmartPresence DataCapture mobile application. The purpose of the facial recognition feature is to recognize the face being recorded is the face that matches the face registration recording pattern for each person. This feature is intended to avoid cheating when doing attendance. Biometrics to date there are three types of fingerprints, facial patterns and eye retina.
SmartPresence uses built-in technology in its application for the recognition of face pattern biometrics. Faces are matched between samples of recorded patterns that have been stored in the database with people’s faces performing attendance taken through the front camera of the smart phone / tablet device. Then in realtime or directly matched to the pattern.
What facial recognition terms are there?
The first thing to do is sample each person’s face in a clean sample state, not blur, light is not excessive or lacking, light must be balanced right and left. If the person concerned using glasses should do the sample without glasses and with glasses.
Face Samples can be more than one. The first face sample has an important role to face face sampling.
What if the next face sample with someone else’s face?
SmartPresence features a new sample comparison feature with an average of samples. the percentage of matches between face samples by default below 70 percent will be rejected.
What if a fake sample of someone else’s face can still enter someone’s sample?
SmartPresence with face biometric recognition stores 2 files when saving patterns. Facial patterns are not only stored in the form of matrix data but also photos. So the sample can still be analyzed further. If there is an inappropriate sample then the sample can be eliminated.
Which areas are recognizable for facial recognition process?
Facial recognition SmartPresence recognizes patterns based on the details of the forehead curve of the eye, the shape of the eye pattern and the midpoint of the eye, the pattern of the nose, the pattern of cheeks, lip patterns and chin patterns.
Can this feature be modified?
SmartPresence provides the basic face setting interface of the face recognition system. Which consists of the first percentage limit of face resemblance, which means if below the percentage then the data record declared rejected or received with a warning.
The second is the confirmation of the face similarity limit which means the lower limit of the absentee record is accepted but with a warning to the system and the upper limit is the percentage of the facial resemblance border.
The third is the limit of face resemblance to the time of registration which means the percentage of the lower limit of the sample average with the new sample being performed. If under that percentage the new face pattern sample will be rejected.
How to know the percentage of face resemblance when doing a record of attendance?
At the time of the absence process on the facial recognition on the mobile application interface DataCapture displayed text that contains a percentage similarity. And this happens instantly. Samples are matched directly to the stored registration pattern.