Multibody Structure-from-Motion in Practice

被引:72
|
作者
Ozden, Kemal Egemen [1 ]
Schindler, Konrad [2 ]
Van Gool, Luc [1 ]
机构
[1] Katholieke Univ Leuven, EAST PSI VISICS, B-3001 Heverlee, Belgium
[2] Tech Univ Darmstadt, Dept Comp Sci, D-64289 Darmstadt, Germany
基金
美国国家科学基金会;
关键词
Structure-from-motion; motion segmentation; scale ambiguity; model selection; affine degeneracy; INDEPENDENTLY MOVING-OBJECTS; STRUCTURE-AND-MOTION; MODEL-SELECTION;
D O I
10.1109/TPAMI.2010.23
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Multibody structure from motion (SfM) is the extension of classical SfM to dynamic scenes with multiple rigidly moving objects. Recent research has unveiled some of the mathematical foundations of the problem, but a practical algorithm which can handle realistic sequences is still missing. In this paper, we discuss the requirements for such an algorithm, highlight theoretical issues and practical problems, and describe how a static structure-from-motion framework needs to be extended to handle real dynamic scenes. Theoretical issues include different situations in which the number of independently moving scene objects changes: Moving objects can enter or leave the field of view, merge into the static background (e. g., when a car is parked), or split off from the background and start moving independently. Practical issues arise due to small freely moving foreground objects with few and short feature tracks. We argue that all of these difficulties need to be handled online as structure-from-motion estimation progresses, and present an exemplary solution using the framework of probabilistic model-scoring.
引用
收藏
页码:1134 / U24
页数:8
相关论文
共 50 条
  • [41] Dorsal stream development in motion and structure-from-motion perception
    Klaver, Peter
    Lichtensteiger, Janine
    Bucher, Kerstin
    Dietrich, Thomas
    Loenneker, Thomas
    Martin, Ernst
    NEUROIMAGE, 2008, 39 (04) : 1815 - 1823
  • [42] The perception of motion and structure in structure-from-motion: Comparisons of affine and Euclidean formulations
    Pollick, FE
    VISION RESEARCH, 1997, 37 (04) : 447 - 466
  • [43] PROJECTIVE STRUCTURE FROM UNCALIBRATED IMAGES - STRUCTURE-FROM-MOTION AND RECOGNITION
    SHASHUA, A
    IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 1994, 16 (08) : 778 - 790
  • [44] Optimal structure-from-motion algorithm for optical flow
    Ohta, N
    Kanatani, K
    IEICE TRANSACTIONS ON INFORMATION AND SYSTEMS, 1995, E78D (12) : 1559 - 1566
  • [45] STRUCTURE-FROM-MOTION INTERPOLATION ACROSS SURFACE DISCONTINUITIES
    SAIDPOUR, A
    BRAUNSTEIN, ML
    HOFFMAN, DD
    INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE, 1992, 33 (04) : 706 - 706
  • [46] Application of Structure-from-Motion photogrammetry to river restoration
    Marteau, Baptiste
    Vericat, Damia
    Gibbins, Chris
    Batalla, Ramon J.
    Green, David R.
    EARTH SURFACE PROCESSES AND LANDFORMS, 2017, 42 (03) : 503 - 515
  • [47] Feature specific segmentation in perceived structure-from-motion
    Li, HCO
    Kingdom, FAA
    VISION RESEARCH, 1999, 39 (05) : 881 - 886
  • [48] Nonrigid structure-from-motion: Estimating shape and motion with hierarchical priors
    Torresani, Lorenzo
    Hertzmann, Aaron
    Bregler, Christoph
    IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2008, 30 (05) : 878 - 892
  • [49] SAR STRUCTURE-FROM-MOTION VIA MATRIX FACTORIZATION
    Hu, Fengyuan
    Jiang, Xue
    Wang, Junfeng
    Liu, Xingzhao
    IGARSS 2023 - 2023 IEEE INTERNATIONAL GEOSCIENCE AND REMOTE SENSING SYMPOSIUM, 2023, : 6967 - 6970
  • [50] The adjacent occluder affects the structure-from-motion perception
    Kitazaki, M.
    PERCEPTION, 2001, 30 : 68 - 68