|
Mjfas templateBog'liq Integration Of Face Recognition Model to a BiometrIntroduction
The era of the Fourth Industrial Revolution has begun. Machines no longer require the aid of humans in order for them to work and
communicate with other machines. These days, machines can just communicate with other machines to execute the tasks they are
designed to do [1]. It is all possible with the help of internet of things. People born in the time of the previous industrial revolution can
only imagine the reach of the technology today.
The Fourth Industrial Revolution bring about the need for better security systems. Since device-to-device communications are
possible and are being developed further at the moment, new security systems or protocols are also being developed. Biometrics are
such security measures developed. Biometrics
centers on the automatic recognition of an individual with the help of the individuals’
distinguishing traits [2].
Facial recognition is by method a part of biometrics. However, without human involvement, machines have a difficulty in differentiating
between one human and another. The solution is to integrate the facial recognition system with an artificial intelligence to help them
‘learn’ the distinguishing features of each individuals’ faces. The artificial intelligence will be trained using machine learning,
specifically the cascade classifier method.
This research aims to create access system with facial recognition by integrating the system with
OpenCV’s pre-trained classifiers
with enhanced training with machine learning. Raspberry Pi will be used to run the facial recognition program. This research will be
limited to creating and training the machine learning model to recognize the features of faces and for the model to be able to
distinguish between individual faces stored in the system.
The paper will discuss the theories used in the research which are cascade classifier and eigenfaces. The hardware Raspberry Pi,
the library OpenCV, the confusion matrix table, and the accuracy, sensitivity and finally the specificity. After discussing theories, the
paper will discuss the methods to create the facial recognition model. Finally, the paper will then discuss the results and analysis from
the research.
|
| |
