Logo LVSN
EnglishAccueil
A proposPersonnesRecherchePublicationsEvenementsProfil
A propos
Publications

 

 

 

 

CERVIM

REPARTI

MIVIM

A new fusion framework for multispectral IR face recognition in the texture space


Moulay Akhloufi and Abdel Hakim Bendada


Abstract - Face recognition is an area of computer vision that has attracted a lot of interest from the research community. A growing demand for robust face recognition software in security applications has driven the development of interesting approaches in this field. A large quantity of research in face recognition deals with visible face images. In the visible spectrum the illumination and face expressions changes represent a significant challenge for the recognition system. To avoid these problems, researchersproposed recently the use of 3D and infrared imaging for face recognition. In this work, we introduce a new framework for multispectral face recognition in the textures space. Active and passive infrared imaging modalities are used and comparison withvisible face recognition is performed. Two multispectral face recognition databases were used in our experiments: Equinox Database (Visible, SWIR, MWIR, LWIR) and Laval University Multispectral Database (Visible, NIR, MWIR, LWIR). The proposed texturespace is based on the use of LBP (Local Binary Pattern) and LTP (Local Ternary Pattern) techniques. Also, a new adaptive texture descriptor is presented. This descriptor uses statistical data extracted from nearest neighbours in order to define the texture spectral kernel. This descriptor called LATP (Local Adaptive Ternary Pattern) is less sensitive to noise in near uniform regions and permit to overcome some of the limitations of LBP and LTP descriptors. The obtained multispectral data are then fused using a multi-scale fusion scheme in order to get interesting face characteristics (figures 4-5). The following multi-scale fusion techniques were used: Laplacian pyramid, FSD pyramid, Ratio pyramid, Contrast pyramid, Gradient pyramid, Morphological pyramid, Discrete wavelets transform (DWT) and Haar wavelets (SIDWT). The multi-scale fusion is performed in different texture channels and between different modes. The obtained results are promising and show a high increase in recognition performance when texture channels are fused in a multi-scale fusion scheme.



Bibtex:

@inproceedings{Akhloufi817,
    author    = { Moulay Akhloufi and Abdel Hakim Bendada },
    title     = { A new fusion framework for multispectral IR face recognition in the texture space },
    booktitle = { Proc. QIRT 10 – Quantitative Infrared Thermography },
    pages     = { X-X },
    address   = { Quebec, QC, Canada },
    year      = { 2010 },
    month     = { July },
    journal   = { QIRT 10 – Quantitative Infrared Thermography },
    language  = { English }
}

Dernière modification: 2010/03/14 par akhloufi

     
   
   

©2002-. Laboratoire de Vision et Systèmes Numériques. Tous droits réservés