Illumination invariant efficient face recognition using a single training image

Bharat Jangid, K.K. Biswas, M. Hanmandlu, Girija CHETTY

    Research output: A Conference proceeding or a Chapter in BookConference contribution

    1 Citation (Scopus)

    Abstract

    This paper presents a single sample face recognition technique which takes care of illumination variations by applying normalization based on Weber's law. Local Directional Pattern (LDP) features are extracted from the normalized face by examining the prominent edge directions at each pixel. The LDP image is divided into non-overlapping windows and each window is treated as a fuzzy set. Treating LDP values as the information source values, entropy features called the information set- based features are extracted from each window. Further, 2DPCA is used to reduce the number of features. These features are augmented with entropy features of the fiducial regions and contour based features for face recognition. A nearest neighbor classifier based on these features is used on Extended Yale B and Face94 datasets and it is shown that compared with other results based on single and multiple training images, the proposed approach results in better recognition accuracy for wide illumination variations in test images. Further the efficiency of the scheme is shown by comparing the number of features needed for recognition.
    Original languageEnglish
    Title of host publicationProceedings International Conference on Digital Image Computing and Applications 2015
    EditorsJamie Sherrah, David Suter
    Place of PublicationAdelaide, Australia
    PublisherIEEE, Institute of Electrical and Electronics Engineers
    Pages1-7
    Number of pages7
    Volume1
    ISBN (Electronic)9781467367943
    ISBN (Print)9781467367950
    DOIs
    Publication statusPublished - 23 Nov 2015
    Event2015 Conference on Digital Image Computing, Techniques and Applications - Adelaide, Adelaide, Australia
    Duration: 23 Nov 201525 Nov 2015

    Conference

    Conference2015 Conference on Digital Image Computing, Techniques and Applications
    Abbreviated titleDICTA 2015
    CountryAustralia
    CityAdelaide
    Period23/11/1525/11/15
    OtherThe International Conference on Digital Image Computing: Techniques and Applications (DICTA) is the main Australian Conference on computer vision, image processing, pattern recognition, and related areas. DICTA was established in 1991 as the premier conference of the Australian Pattern Recognition Society (APRS). DICTA 2015 is endorsed by the IAPR and technically co-sponsored by the IEEE

    Fingerprint

    Face recognition
    Entropy
    Lighting
    Fuzzy sets
    Classifiers
    Pixels

    Cite this

    Jangid, B., Biswas, K. K., Hanmandlu, M., & CHETTY, G. (2015). Illumination invariant efficient face recognition using a single training image. In J. Sherrah, & D. Suter (Eds.), Proceedings International Conference on Digital Image Computing and Applications 2015 (Vol. 1, pp. 1-7). Adelaide, Australia: IEEE, Institute of Electrical and Electronics Engineers. https://doi.org/10.1109/DICTA.2015.7371266
    Jangid, Bharat ; Biswas, K.K. ; Hanmandlu, M. ; CHETTY, Girija. / Illumination invariant efficient face recognition using a single training image. Proceedings International Conference on Digital Image Computing and Applications 2015. editor / Jamie Sherrah ; David Suter. Vol. 1 Adelaide, Australia : IEEE, Institute of Electrical and Electronics Engineers, 2015. pp. 1-7
    @inproceedings{f0a39a2906e646648b9457d0dd4cbcea,
    title = "Illumination invariant efficient face recognition using a single training image",
    abstract = "This paper presents a single sample face recognition technique which takes care of illumination variations by applying normalization based on Weber's law. Local Directional Pattern (LDP) features are extracted from the normalized face by examining the prominent edge directions at each pixel. The LDP image is divided into non-overlapping windows and each window is treated as a fuzzy set. Treating LDP values as the information source values, entropy features called the information set- based features are extracted from each window. Further, 2DPCA is used to reduce the number of features. These features are augmented with entropy features of the fiducial regions and contour based features for face recognition. A nearest neighbor classifier based on these features is used on Extended Yale B and Face94 datasets and it is shown that compared with other results based on single and multiple training images, the proposed approach results in better recognition accuracy for wide illumination variations in test images. Further the efficiency of the scheme is shown by comparing the number of features needed for recognition.",
    keywords = "illumination-invariant, LDP-features, Entropy-function",
    author = "Bharat Jangid and K.K. Biswas and M. Hanmandlu and Girija CHETTY",
    year = "2015",
    month = "11",
    day = "23",
    doi = "10.1109/DICTA.2015.7371266",
    language = "English",
    isbn = "9781467367950",
    volume = "1",
    pages = "1--7",
    editor = "Jamie Sherrah and David Suter",
    booktitle = "Proceedings International Conference on Digital Image Computing and Applications 2015",
    publisher = "IEEE, Institute of Electrical and Electronics Engineers",
    address = "United States",

    }

    Jangid, B, Biswas, KK, Hanmandlu, M & CHETTY, G 2015, Illumination invariant efficient face recognition using a single training image. in J Sherrah & D Suter (eds), Proceedings International Conference on Digital Image Computing and Applications 2015. vol. 1, IEEE, Institute of Electrical and Electronics Engineers, Adelaide, Australia, pp. 1-7, 2015 Conference on Digital Image Computing, Techniques and Applications, Adelaide, Australia, 23/11/15. https://doi.org/10.1109/DICTA.2015.7371266

    Illumination invariant efficient face recognition using a single training image. / Jangid, Bharat; Biswas, K.K.; Hanmandlu, M.; CHETTY, Girija.

    Proceedings International Conference on Digital Image Computing and Applications 2015. ed. / Jamie Sherrah; David Suter. Vol. 1 Adelaide, Australia : IEEE, Institute of Electrical and Electronics Engineers, 2015. p. 1-7.

    Research output: A Conference proceeding or a Chapter in BookConference contribution

    TY - GEN

    T1 - Illumination invariant efficient face recognition using a single training image

    AU - Jangid, Bharat

    AU - Biswas, K.K.

    AU - Hanmandlu, M.

    AU - CHETTY, Girija

    PY - 2015/11/23

    Y1 - 2015/11/23

    N2 - This paper presents a single sample face recognition technique which takes care of illumination variations by applying normalization based on Weber's law. Local Directional Pattern (LDP) features are extracted from the normalized face by examining the prominent edge directions at each pixel. The LDP image is divided into non-overlapping windows and each window is treated as a fuzzy set. Treating LDP values as the information source values, entropy features called the information set- based features are extracted from each window. Further, 2DPCA is used to reduce the number of features. These features are augmented with entropy features of the fiducial regions and contour based features for face recognition. A nearest neighbor classifier based on these features is used on Extended Yale B and Face94 datasets and it is shown that compared with other results based on single and multiple training images, the proposed approach results in better recognition accuracy for wide illumination variations in test images. Further the efficiency of the scheme is shown by comparing the number of features needed for recognition.

    AB - This paper presents a single sample face recognition technique which takes care of illumination variations by applying normalization based on Weber's law. Local Directional Pattern (LDP) features are extracted from the normalized face by examining the prominent edge directions at each pixel. The LDP image is divided into non-overlapping windows and each window is treated as a fuzzy set. Treating LDP values as the information source values, entropy features called the information set- based features are extracted from each window. Further, 2DPCA is used to reduce the number of features. These features are augmented with entropy features of the fiducial regions and contour based features for face recognition. A nearest neighbor classifier based on these features is used on Extended Yale B and Face94 datasets and it is shown that compared with other results based on single and multiple training images, the proposed approach results in better recognition accuracy for wide illumination variations in test images. Further the efficiency of the scheme is shown by comparing the number of features needed for recognition.

    KW - illumination-invariant

    KW - LDP-features

    KW - Entropy-function

    U2 - 10.1109/DICTA.2015.7371266

    DO - 10.1109/DICTA.2015.7371266

    M3 - Conference contribution

    SN - 9781467367950

    VL - 1

    SP - 1

    EP - 7

    BT - Proceedings International Conference on Digital Image Computing and Applications 2015

    A2 - Sherrah, Jamie

    A2 - Suter, David

    PB - IEEE, Institute of Electrical and Electronics Engineers

    CY - Adelaide, Australia

    ER -

    Jangid B, Biswas KK, Hanmandlu M, CHETTY G. Illumination invariant efficient face recognition using a single training image. In Sherrah J, Suter D, editors, Proceedings International Conference on Digital Image Computing and Applications 2015. Vol. 1. Adelaide, Australia: IEEE, Institute of Electrical and Electronics Engineers. 2015. p. 1-7 https://doi.org/10.1109/DICTA.2015.7371266