Yazar "Abeysundera, Hasith Pasindu" seçeneğine göre listele

Listeleniyor 1 - 3 / 3
  • Küçük Resim
    Yayın
    Nearest neighbor weighted average customization for modeling faces
    (Springer, 2013-10) Abeysundera, Hasith Pasindu; Benli, Kristin Surpuhi; Eskil, Mustafa Taner
    In this paper, we present an anatomically accurate generic wireframe face model and an efficient customization method for modeling human faces. We use a single 2D image for customization of the generic model. We employ perspective projection to estimate 3D coordinates of the 2D facial landmarks in the image. The non-landmark vertices of the 3D model are shifted using the translations of k nearest landmark vertices, inversely weighted by the square of their distances. We demonstrate on Photoface and Bosphorus 3D face data sets that the proposed method achieves substantially low relative error values with modest time complexity.
  • Küçük Resim
    Yayın
    Palmprint verification using SIFT majority voting
    (Springer-Verlag, 2012) Abeysundera, Hasith Pasindu; Eskil, Mustafa Taner
    In this paper we illustrate the implementation of a robust, real-time biometric system for identity verification based on palmprint images. The palmprint images are preprocessed to align the major axes of hand shapes and to extract the palm region. We extract features using Scale Invariant Feature Transform (SIFT). Classification of individual SIFT features is done through KNN. The class of the hand image is decided by a majority based voting among its classified SIFT features. We demonstrate on the CASIA and PolyU datasets that the proposed system achieves authentication accuracy comparable to other state of the art algorithms.
  • Küçük Resim
    Yayın
    Semi-automatic customization for modeling human face
    (Işık Üniversitesi, 2012-02-14) Abeysundera, Hasith Pasindu; Eskil, Mustafa Taner; Işık Üniversitesi, Fen Bilimleri Enstitüsü, Bilgisayar Mühendisliği Yüksek Lisans Programı
    Model-based vision has firmly established its roots as a robust approach in rec-ognizing and locating known traits in rigid objects even under the presence of noise, clutter and occlusion. However the application of such systems has not displayed the same efficiency in modeling non-rigid objects. The dilemma with the prevailing modeling techniques is that that they compensate model specificity to accommodate variability, or the vice versa compromising the robustness of the 3 dimensional model during the image interpretation progression. Face, being a non rigid and a sophisticated structure makes it more arduous to model, using such approaches. In this study we have presented a novel method in modeling 3 dimensional images employing a generic wireframe and a single 2 dimensional image. Known traits are located in the 3 dimensional space using a variant of ray tracing method. Non-landmark traits are positioned employing a nearest neighbor weighted average customization. Proposed technique has proven its robustness in the experiments conducted employing the Bosphorous database. Furthermore the relative error values attained employing NNWA customization illustrated significantly low val»ues. We compared the obtained results with the ASM and Procrustes Analysis.