Advances in Computer Science and Information Technology. Networks and Communications. Second International Conference, CCSIT 2012, Bangalore, India, January 2-4, 2012. Proceedings, Part I

Research Article

Weighted Angle Based Approach for Face Recognition

Download
242 downloads
  • @INPROCEEDINGS{10.1007/978-3-642-27299-8_34,
        author={M. Koteswara Rao and K. Veeraswamy and K. Anitha Sheela and B. Chandra Mohan},
        title={Weighted Angle Based Approach for Face Recognition},
        proceedings={Advances in Computer Science and Information Technology. Networks and Communications. Second International Conference, CCSIT 2012, Bangalore, India, January 2-4, 2012. Proceedings, Part I},
        proceedings_a={CCSIT PART I},
        year={2012},
        month={11},
        keywords={Sub-pattern Eigenvectors Weighted angle},
        doi={10.1007/978-3-642-27299-8_34}
    }
    
  • M. Koteswara Rao
    K. Veeraswamy
    K. Anitha Sheela
    B. Chandra Mohan
    Year: 2012
    Weighted Angle Based Approach for Face Recognition
    CCSIT PART I
    Springer
    DOI: 10.1007/978-3-642-27299-8_34
M. Koteswara Rao1,*, K. Veeraswamy2,*, K. Anitha Sheela3,*, B. Chandra Mohan4,*
  • 1: QIS College of Engineering & Technology
  • 2: Qis College of Engineering & Technology
  • 3: JNTUH
  • 4: Bapatla Engineering College
*Contact email: koteshproject@gmail.com, kilarivs@yahoo.com, kanithasheela@gmail.com, chadrabhuma@gmail.com

Abstract

A Face recognition scheme using weighted angle based approach is proposed in this paper. In content based image retrieval, Face recognition system performs fast and accurate detection from database. Feature vector based on Eigen vectors of sub images is used for recognition. Image is partitioned into sub images. Sub parts are rearranged into rows and column matrices. Eigenvectors are computed for these matrices. Global feature vector is generated and weighted angle distance is used for face recognition. Experiments performed on benchmark face database (YALE) indicated that the proposed weighted angle based approach has better recognition performance in terms of average recognized rate and retrieval time compared to the existing methods.