Accurate 3D Reconstruction of Dynamic Scenes from Monocular Image Sequences with Severe Occlusions

被引:3
|
作者
Golyanik, Vladislav [1 ]
Fetzer, Torben
Stricker, Didier
机构
[1] Univ Kaiserslautern, Dept Comp Sci, Kaiserslautern, Germany
关键词
STRUCTURE-FROM-MOTION; SHAPE;
D O I
10.1109/WACV.2017.38
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
The paper introduces an accurate solution to dense orthographic Non-Rigid Structure from Motion (NRSfM) in scenarios with severe occlusions or, likewise, inaccurate correspondences. We integrate a shape prior term into variational optimisation framework. It allows to penalize irregularities of the time-varying structure on the per-pixel level if correspondence quality indicator such as an occlusion tensor is available. We make a realistic assumption that several non-occluded views of the scene are sufficient to estimate an initial shape prior, though the entire observed scene may exhibit non-rigid deformations. Experiments on synthetic and real image data show that the proposed framework significantly outperforms state of the art methods for correspondence establishment in combination with the state of the art NRSfM methods. Together with the profound insights into optimisation methods, implementation details for heterogeneous platforms are provided.
引用
收藏
页码:282 / 291
页数:10
相关论文
共 50 条
  • [21] Total Variation Based 3D Reconstruction from Monocular Laparoscopic Sequences
    Marcinczak, Jan Marek
    Grigat, Rolf-Rainer
    ABDOMINAL IMAGING: COMPUTATIONAL AND CLINICAL APPLICATIONS, 2014, 8676 : 239 - 247
  • [22] Bayesian reconstruction of 3D shapes and scenes from a single image
    Han, F
    Zhu, SC
    FIRST IEEE INTERNATIONAL WORKSHOP ON HIGHER-LEVEL KNOWLEDGE IN 3D MODELING AND MOTION ANALYSIS, PROCEEDINGS, 2003, : 12 - 20
  • [23] 3D model localisation using high-resolution reconstruction of monocular image sequences
    Serra, B
    AUTOMATIC TARGET RECOGNITION VII, 1997, 3069 : 282 - 293
  • [24] A Dense Pipeline for 3D Reconstruction from Image Sequences
    Schneevoigt, Timm
    Schroers, Christopher
    Weickert, Joachim
    PATTERN RECOGNITION, GCPR 2014, 2014, 8753 : 629 - 640
  • [25] 3D Reconstruction of Indoor Scenes Using RGB-D Monocular Vision
    Liu, Sanmao
    Zhu, Wenqiu
    Zhang, Canqing
    Sun, Wenjing
    2016 INTERNATIONAL CONFERENCE ON ROBOTS & INTELLIGENT SYSTEM (ICRIS), 2016, : 1 - 7
  • [26] 3D reconstruction of dynamic scenes dedicated to an image sensor of a visual intracortical stimulator
    Doljanu, A. D.
    Sawan, M.
    2006 INTERNATIONAL WORKSHOP ON COMPUTER ARCHITECTURE FOR MACHINE PERCEPTION AND SENSING, 2006, : 177 - 178
  • [27] Dense Reconstruction of Underwater Scenes from Monocular Sequences of Images
    Rzhanov, Yuri
    Hu, Han
    Boyer, Thierry
    OCEANS 2014 - TAIPEI, 2014,
  • [28] 3D Reconstruction of a Smooth Articulated Trajectory from a Monocular Image Sequence
    Park, Hyun Soo
    Sheikh, Yaser
    2011 IEEE INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV), 2011, : 201 - 208
  • [29] 3D reconstruction of human skeleton from single images or monocular video sequences
    Remondino, F
    Roditakis, A
    PATTERN RECOGNITION, PROCEEDINGS, 2003, 2781 : 100 - 107
  • [30] Non-linear statistical models for the 3D reconstruction of human pose and motion from monocular image sequences
    Bowden, R
    Mitchell, TA
    Sarhadi, M
    IMAGE AND VISION COMPUTING, 2000, 18 (09) : 729 - 737