An Efficient Face Recognition System Based on Sub-Window Extraction Algorithm
Manish Gupta1, Govind Sharma2

Manuscript received on December 07, 2011. | Revised Manuscript received on December 23, 2011. | Manuscript published on January 05, 2012. | PP: 304-310 | Volume-1 Issue-6, January 2012. | Retrieval Number: F0333121611/2012©BEIESP
Open Access | Ethics and Policies | Cite
© The Authors. Published By: Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: In this paper, an efficient face recognition system based on sub-window extraction algorithm and recognition based on principal component analysis (PCA) and Back propagation algorithm is proposed. Our proposed method works on two phases: Extraction phase and Recognition phase. In extraction phase, face images are captured from different sources and then enhanced using filtering, clipping and histogram equalization. Enhanced images are converted into edge images using Sobel operator and then converted into binary images. Finally sub windows from extracted using proposed sub windows extraction algorithm and extract different features (mouth, eyes, nose etc.) from these sub windows. In recognition phase, back propagation algorithm (BPA) and PCA algorithm is used. The experiments are carried out using IIIM_Gwalior database, IIT Kanpur database and Face_94 database.
Keywords: Sub-windows extraction, principal component analysis (PCA), Back propagation algorithm (BPA), Face recognition, Neural Network.