Being able to recover the shape of 3D deformable surfaces from a single video stream would make it possible to field reconstruction systems that run on widely available hardware without requiring specialized devices. However, because many different 3D shapes can have virtually the same projection, such monocular shape recovery is inherently ambiguous. In this survey, we will review the two main classes of techniques that have proved most effective so far: The template-based methods that rely on establishing correspondences with a reference image in which the shape is already known, and non-rigid structure-from-motion techniques that exploit points tracked across the sequences to reconstruct a completely unknown shape. In both cases, we will formalize the approach, discuss its inherent ambiguities, and present the practical solutions that have been proposed to resolve them. To conclude, we will suggest directions for future research

    2. Early approaches to non-rigid reconstruction -- Physics-based models -- The finite element method -- Physics-based methods for computer graphics -- Physics-based methods for computer vision -- Learned deformation models -- Statistical learning methods -- Learned models for non-rigid modeling -- Regularization via shape parameterization -- Legacy of the previous approaches --

    1. Introduction --

    3. Formalizing template-based reconstruction -- Problem definition -- Motivation -- Camera models -- 3D-to-2D correspondences -- Linear formulation -- Ambiguities under weak perspective projection -- Ambiguities under full perspective projection --

    4. Performing template-based reconstruction -- Imposing temporal consistency -- Imposing geometric constraints -- Developable surfaces -- Smooth surfaces -- Distance constraints --

    5. Formalizing non-rigid structure from motion -- Problem definition -- NRSFM under weak perspective projection -- NRSFM under full perspective projection -- Ambiguities of NRSFM -- The missing data problem --

    6. Performing non-rigid structure from motion -- Orthonormality constraints -- Imposing temporal consistency -- From basis shapes to basis trajectories -- Imposing geometric constraints -- Global constraints -- Local constraints -- Splitting a global surface into local ones --

    7. Future directions -- Bibliography -- Authors' biographies


    Zugriff

    Zugriff über TIB

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Deformable Surface 3D Reconstruction from Monocular Images


    Beteiligte:
    Salzmann, Matthieu (Autor:in) / Fua, Pascal (Autor:in)


    Erscheinungsdatum :

    2010


    Format / Umfang :

    1 Online-Ressource (113 Seiten)


    Anmerkungen:

    Campusweiter Zugriff (Universität Hannover) - Vervielfältigungen (z.B. Kopien, Downloads) sind nur von einzelnen Kapiteln oder Seiten und nur zum eigenen wissenschaftlichen Gebrauch erlaubt. Keine Weitergabe an Dritte. Kein systematisches Downloaden durch Robots.
    Description based upon print version of record




    Medientyp :

    Buch


    Format :

    Elektronische Ressource


    Sprache :

    Englisch



    Klassifikation :


    Robust deformable shape reconstruction from monocular video with manifold forests

    Tao, L. / Matuszewski, B. J. | British Library Online Contents | 2016


    Monocular 3D Object Detection Utilizing Auxiliary Learning With Deformable Convolution

    Chen, Jiun-Han / Shieh, Jeng-Lun / Haq, Muhamad Amirul et al. | IEEE | 2024



    3D Object Detection from Consecutive Monocular Images

    Cheng, Chia-Chun / Lai, Shang-Hong | British Library Conference Proceedings | 2021