All Volumes & Issues

Volume 6, Issue 1, March 2019

The Novel Approach towards Face Detection by Corner Detection Using Special Morphological Masking System and Fast Algorithm

Tumpa Dey Dibyendu Ghoshal

  • Abstract

    In this work we have performed a face recognition technique using corner detection algorithm. Here we have used fast corner detection algorithm but modified it by applying a masking technique to make a robust technique with respect to noise. This paper describes twelve facial landmarks which include left eye corners, right eye corners, left eyebrows, right eyebrows, lip corners, nostrils. It consists of two parts; in the first step we have done the masking technique to filter the image from noise. In the second step we applied the fast corner detection algorithm to detect the facial landmarks. The Fast corner detector works on the corner response function (CRF), which is computed as a minimum change of intensity over all possible direction. Lastly we have done a comparison study with other filtering techniques with respect to proposed masking technique. In this paper, we have done the experiments using IRIS face Database, BioID and Cohn Kanade Database. Recognition rate achieved by applying the proposed method are very well.

  • Keywords

    Face Recognition; facial landmarks; fast corner detection; masking technique; corner response function; IRIS face database; BioID database; Cohn kanade database

  • Author Affiliations

    Tumpa Dey 1

    Dibyendu Ghoshal 2

    1. M.B.B College, Agartala,India
    2. NIT Agartala, Agartala,India
  • Dates

    Manuscript received : 13 December 2018
    Manuscript revised   : 16 January 2019
    Accepted                     : 21 February 2019

  • Cite this article as:
    Tumpa Dey, Dibyendu Ghoshal, Int.j.of IJACCS(2019)


    Print ISSN           : 2347 - 9280
    Online ISSN        : 2347 - 9299
    Publisher Name  : Sri Eshwar Publications, Coimbatore, Tamilnadu, India.

    Article Informations

    Information for Authors

    Journal Services

    Abstracted/Indexed in