An effective model for Heterogeneous Face recognition in a static background

  • Unique Paper ID: 153486
  • Volume: 8
  • Issue: 7
  • PageNo: 372-376
  • Abstract:
  • Real time faces images are captured in different spectral bands are referred to as heterogeneous. In this paper, the new approach based on low resolution heterogeneous face recognition and synthesis is proposed. In the recognition section, multi block local binary pattern (MBLBP) is used as facial representation for low resolution images. Then Canonical Correlation Analysis (CCA) is a applied to learn the mapping between the different LBP-face patterns. The corresponding matching scores are calculated in the CCA subspace for the final decision. In the synthesis section, according to the CCA transformation matrices obtained above, we apply ridge regression to determine an approximate linear relation between the target pattern image and the projection vector of the probe. The work shows a practical solution for reliable heterogeneous face synthesis.

Cite This Article

  • ISSN: 2349-6002
  • Volume: 8
  • Issue: 7
  • PageNo: 372-376

An effective model for Heterogeneous Face recognition in a static background

Related Articles