Implementation of Embedded Face Detection System on Raspberry Pi Platform

Abstract

Face detection is one of the biometric identification methods, which identifies or verifies a person from a digital image or a video frame containing a person face. Face recognition is recently popular in security system as a tool for identifying a suspected person or access control. A traditional facial recognition algorithm, which identifies facial features by extracting features (i.e., shape of eyes, nose, and jaw) of the subject, is implemented in computing platform capable of processing an image or a video frame at decent speed. The extracted feature is then compared with the existing features stored in the database for finding matched features. The position of the subject image is crucial to the accuracy of facial recognition algorithm. A stationary platform may have high computing power, but the position of the subject as images cannot be controlled. In contrast, a mobile platform can position itself in order to obtain better subject images. This paper presents the implementation of face recognition algorithm on the embedded system, which can be installed on a mobile platform in order to overcome the limitation of the position of subject as image. In this system, Raspberry Pi based on Linux operation system is used as an embedded system. The facial recognition algorithm is developed using python programming language in order to extract the positions and shapes of eyes, nose, and jaw of the subject as face. The system connects to the database via WiFi and searches for matching features.