Coarse-to-fine stereo vision with accurate 3D boundaries

被引:32
|
作者
Sizintsev, Mikhail [1 ]
Wildes, Richard P.
机构
[1] York Univ, Dept Comp Sci & Engn, Toronto, ON M3J 1P3, Canada
关键词
Computer vision; Stereo; Coarse-to-fine; Occlusions; Real-time algorithms; ENERGY MINIMIZATION; ALGORITHM; PROPAGATION;
D O I
10.1016/j.imavis.2009.06.008
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
This paper presents methods for efficient recovery of accurate binocular disparity estimates in the vicinity of 3D surface discontinuities. Of particular concern are methods that impact coarse-to-fine, local block-based matching as it forms the basis of the fastest and the most resource efficient stereo computation procedures. A novel coarse-to-fine refinement procedure that adapts match window support across scale to ameliorate corruption of disparity estimates near boundaries is presented. Extensions are included to account for half-occlusions and colour uniformity. Empirical results show that incorporation of these advances in the standard coarse-to-fine, block matching framework reduces disparity errors by more than a factor of two, while performing little extra computation, preserving low complexity and the parallel/pipeline nature of the framework. Moreover, the proposed advances prove to be beneficial for CTF global matchers as well. (C) 2009 Elsevier B.V. All rights reserved.
引用
收藏
页码:352 / 366
页数:15
相关论文
共 50 条
  • [21] Coarse-to-fine information integration in human vision
    Petras, Kirsten
    ten Oever, Sanne
    Jacobs, Christianne
    Goffaux, Valerie
    NEUROIMAGE, 2019, 186 : 103 - 112
  • [22] Coarse-to-fine 3D road model registration for traffic video augmentation
    Cui, Zhichao
    Li, Yaochen
    Zhang, Chi
    Liu, Yuehu
    Ren, Fuji
    IET IMAGE PROCESSING, 2020, 14 (12) : 2690 - 2700
  • [23] Photon-Efficient 3D Reconstruction with A Coarse-to-Fine Neural Network
    Guo, Shangwei
    Lai, Zhengchao
    Li, Jun
    Han, Shaokun
    OPTICS AND LASERS IN ENGINEERING, 2022, 159
  • [24] TransPillars: Coarse-to-Fine Aggregation for Multi-Frame 3D Object Detection
    Luo, Zhipeng
    Zhang, Gongjie
    Zhou, Changqing
    Liu, Tianrui
    Lu, Shijian
    Pan, Liang
    2023 IEEE/CVF WINTER CONFERENCE ON APPLICATIONS OF COMPUTER VISION (WACV), 2023, : 4219 - 4228
  • [25] A Coarse-to-Fine Approach for Rock Bolt Detection From 3D Point Clouds
    Saydam, Sarp
    Liu, Boge
    Li, Binghao
    Zhang, Wenjie
    Singh, Sarvesh Kumar
    Raval, Simit
    IEEE ACCESS, 2021, 9 : 148873 - 148883
  • [26] A Coarse-to-Fine Approach for 3D Facial Landmarking by Using Deep Feature Fusion
    Wang, Kai
    Zhao, Xi
    Gao, Wanshun
    Zou, Jianhua
    SYMMETRY-BASEL, 2018, 10 (08):
  • [27] Coarse-to-fine multiview 3d face reconstruction using multiple geometrical features
    Dai, Peng
    Wang, Xue
    Zhang, Weihang
    MULTIMEDIA TOOLS AND APPLICATIONS, 2018, 77 (01) : 939 - 966
  • [28] Vehicle trajectory clustering based on 3D information via a coarse-to-fine strategy
    Huansheng Song
    Xuan Wang
    Cui Hua
    Weixing Wang
    Qi Guan
    Zhaoyang Zhang
    Soft Computing, 2018, 22 : 1433 - 1444
  • [29] 3D mouse brain reconstruction from histology using a coarse-to-fine approach
    Yushkevich, Paul A.
    Avants, Brian B.
    Ng, Lydia
    Hawrylycz, Michael
    Burstein, Pablo D.
    Zhang, Hui
    Gee, James C.
    BIOMEDICAL IMAGE REGISTRATION, PROCEEDINGS, 2006, 4057 : 230 - 237
  • [30] A Novel Coarse-to-fine Registration for 3D Point Cloud Based on Feature Points
    Huo, Guan-ying
    Jiang, Xin
    Ye, Dan-lei
    Su, Cheng
    Lu, Ze-hong
    Wang, Bo-lun
    Zheng, Zhi-ming
    2018 3RD INTERNATIONAL CONFERENCE ON COMPUTATIONAL MODELING, SIMULATION AND APPLIED MATHEMATICS (CMSAM 2018), 2018, 310 : 385 - 392