Ensemble Face Recognition System Using Dense Local Graph Structure

This paper presents an ensemble face recognition system which makes use of a novel Dense Local Graph Structure (D-LGS). This descriptor uses additional graph structure along with its own local graph structure. This additional local graph structure is generated from original symmetric LGS by finding additional corner pixels through bilinear interpolation of neighbourhood pixels. These corner pixels lead to most stable features and information related to local deformation of the image. In this proposed ensemble system, three classifiers namely K-NN, Chi-square and correlation coefficient are used. Further the proposed approach fuses the decisions of individual classifiers through OR rule, majority voting and AND rule. To evaluate the performance of proposed ensemble system the experiment is conducted with two face databases viz. the AT&T (formerly the ORL database) and the UFI (Unconstrained Facial Image) database. The experimental results exhibit considerable improvement of the ensemble face recognition system on the use of novel descriptor.