A Review and Construction of a Real-time Facial Recognition System
Eduardo H. Teixeira, Samuel Baraldi Mafra, Joel J. P. C. Rodrigues, Werner Augusto A. N. da Silveira, Ousmane Noël Diallo · 2020
The evolution of surveillance technologies allows a reduction in human interaction with the process, since most of the monitoring functions performed by an individual can be replaced by detection and recognition techniques in real-time. This paper proposes the development of a surveillance system, which uses these techniques to identify individuals present within the field of view of camera. A combination of the Histogram of Oriented Gradient and Support Vector Machine techniques is applied for face detection, while a Residual Network is used during the stage of recognizing individuals. This shows the possibility of implementing this set of techniques, even in hardware with processing limitations.