An approach to real-time portable device for face recognition system

Abstract

Face is the most visible organ of human which can be used to differentiate one person from another. Embedded face recognition is quite a challenging era of biometric which can be used in different applications. There are currently commercially available systems for face recognition, but they are bulky, expensive, and proprietary. In this paper we will discuss about a real-time, portable, low cost face recognition system, developed by us, we named it as RPIPD-v1 (Real Time Portable Image Processing Device Version 1) which can work as a standalone system. This system can detect and recognize human faces and there is no such portable device that is capable of having this feature together. Our main aim was to glue these two tasks (Face Detection & Face Recognition) and build a system with the hardware and software to demonstrate the performance. The accuracy of our device is about 60% in distinguishing known and unknown faces. This is a general output of our device for critical situations. For the best performance of recognizing a face we need to highlight the individuals face with proper lighting and the individual's face must be in front of the camera properly. This device is made totally portable and standalone. Small size and low power enables this device to be placed in unique environments, collecting images autonomously for face recognition and other image processing applications. Low cost is another advantage which will allow this device to be used in various places as an effective solution.