We propose a new family of perfect reconstruction, non-redundant, and multiresolution geometrical image transforms using the wavelet transform in conjunction with modified versions of directional filter banks (DFB). In the proposed versions of DFB, we use either horizontal or vertical directional decomposition. Taking advantage of the wavelet transform that has efficient nonlinear approximation property, we add the important feature of directionality by applying the modified and regular DFB to the subbands of a few finest wavelet levels. This way we can eliminate a major portion of the artifacts usually introduced when DFB are used. The proposed hybrid wavelets and DFB (HWD) transform family provides visual and PSNR improvements over the wavelet and contourlet transforms.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    New image transforms using hybrid wavelets and directional filter banks: analysis and design


    Contributors:
    Eslami, R. (author) / Radha, H. (author)


    Publication date :

    2005-01-01


    Size :

    705866 byte




    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English



    New Image Transforms using Hybrid Wavelets and Directional Filter Banks: Analysis and Design

    Eslami, R. / Radha, H. | British Library Conference Proceedings | 2005


    Causal stable matched IIR wavelets and perfect reconstruction filter banks [4056-28]

    Rao, R. M. / Lopez-Estrada, A. A. / Bopardikar, A. S. et al. | British Library Conference Proceedings | 2000


    On Symmetric Extensions, Orthogonal Transforms of Images, and Paraunitary Filter Banks

    De Queiroz, R. L. / Rao, K. R. / IEEE; Signal Processing Society | British Library Conference Proceedings | 1994



    Efficient Image Scrambling using Polyphase Filter Banks

    Creusere, C. D. / Mitra, S. K. / IEEE; Signal Processing Society | British Library Conference Proceedings | 1994