Improving Face Recognition Performance Using Similarity Feature-Based Selection and Classification Algorithm

• Published in: 2013 Second International Conference on Robot, Vision and Signal Processing pp. 56 - 60
• Main Authors: Chi-Kien Tran, Tsair-Fwu Lee, Chiu-Ching Tuan, Chi-Heng Lu, Pei-Ju Chao
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01-12-2013
• Subjects: Classification algorithms; Euclidean distance; Face; Face recognition; Feature extraction; histogram; Histograms; Manhattan distance; pixel values; similarity feature; Training; Vectors
• ISSN: 2376-9793; 2376-9807
• DOI: 10.1109/RVSP.2013.21

In this paper, we propose the effective similarity feature-based selection and classification algorithm to select similarity features on the training images and to classify face images in face recognition system. The experiments are conducted on The ORL Database of Faces, which consists of 400 images of 40 individuals. Two face recognition systems, one based on the histogram-based feature, and the other based on the feature which is the mean of pixel values in window with size of 4×4 (M4×4), are developed. Euclidean distance and Manhattan distance are taken as distance metrics for the classification method. The results indicated that the proposed algorithms not only reduce the dimensions of feature space, but also achieve a mean recognition accuracy that is 1.55% ± 11.31% better compared to conventional algorithms.