Embedded Vision System for Autonomous Face Recognizer Robot using Eigenface Detector-PCA Improved Face Detection for Largely Occluded Faces

Abstract

The problem is to propose a method for detecting the face and recognizing the face for a robot. This is a challenging problem because faces are non-rigid and have a high degree of variability in size, shape, colour, and texture. In this report, an embedded vision system is built to locate, recognize and track the faces. Three individual sub-systems are present. The first one is face detection based on Haar-like classifier and ViolaJones object detector. The second one is face recognition system based on Eigenface detector using principal component analysis method. The third one is face tracking based on Haar Cascade Classifier. For fast and an efficient embedded vision system for an autonomous robot, the software used is OpenCV library. The system developed has improved performance and can also detect largely occluded faces.