Title: Face recognition using multiple content-based image features for biometric security applications
Authors: Madeena Sultana; Marina L. Gavrilova
Addresses: Department of Computer Science, University of Calgary, 2500 University Drive NW, Calgary, AB, Canada ' Department of Computer Science, University of Calgary, 2500 University Drive NW, Calgary, AB, Canada
Abstract: During the era of internet, content-based image retrieval (CBIR) systems, where images are searched based on their visual contents, have an increasing demand for numerous real world applications. However, the potential of using multiple CBIR-based features for biometric recognition remains largely unexplored. This research presents an in-depth analysis of current research trends of CBIR and its potential applications in the field of biometric security. A novel content-based face recognition system is proposed and experimental results are provided to strengthen the material of this article. In the proposed face recognition system, three content-based low level features: colour, texture, and shape are combined to enhance the recognition accuracy. Moreover, the simplicity and ease of computation of the exploited methods reduce computation time. Experimental results show that the proposed multiple low level feature-based method outperforms single feature-based face recognition systems.
Keywords: face recognition; content-based image retrieval; CBIR; colour histogram; Gabor filter; histogram intersection; affine moment invariants; AMI; pseudo Zernike moment invariants; image features; biometric security; biometrics.
International Journal of Biometrics, 2014 Vol.6 No.4, pp.414 - 434
Available online: 27 Jan 2015 *Full-text access for editors Access for subscribers Purchase this article Comment on this article