Home > Published Issues > 2015 > Volume 3, No. 1, June 2015 >

Face Recognition Using String Grammar Nearest Neighbor Technique

P. Kasemsumran 1, S. Auephanwiriyakul 1,2, and N. Theera-Umpon 2,3
1. Computer Engineering Department, Faculty of Engineering, Chiang Mai University, Chiang Mai, Thailand
2. Biomedical Engineering Center, Chiang Mai University, Chiang Mai, Thailand
3. Electrical Engineering Department, Faculty of Engineering, Chiang Mai University, Chiang Mai, Thailand

Abstract—Face recognition has become one of the important biometrics in many applications. However, there is a problem of collecting more than one image per person in the training data set, the so-called “one sample per person problem”. Hence in this paper, we develop a face recognition system with a string grammar nearest neighbor (sgNN) to cope with the problem. We implement our system in three data sets, i.e., ORL, MIT-CBCL, and Georgia Tech databases. The recognition rates of the test data set from three databases are 88.25%, 87.50%, and 70.71%, respectively.

Index Terms—face recognition, one sample per person, string grammar, nearest neighbor, Levenshtein distance

Cite: P. Kasemsumran, S. Auephanwiriyakul, and N. Theera-Umpon, "Face Recognition Using String Grammar Nearest Neighbor Technique," Journal of Image and Graphics, Vol. 3, No. 1, pp. 6-10, June 2015. doi: 10.18178/joig.3.1.6-10

PREVIOUS PAPER
A Comparison of South East Asian Face Emotion Classification Based on Optimized Ellipse Data Using Clustering Technique
NEXT PAPER
Multi-Layer Framed Offline Signature Recognition Algorithm