FEATS: Synthetic Feature Tracks for Structure from Motion Evaluation

被引:8
|
作者
DeGol, Joseph [1 ]
Lee, Jae Yong [1 ]
Kataria, Rajbir [1 ]
Yuan, Daniel [1 ]
Bretl, Timothy [1 ]
Hoiem, Derek [1 ,2 ]
机构
[1] Univ Illinois, Champaign, IL 61820 USA
[2] Reconstruct Inc, Champaign, IL USA
关键词
D O I
10.1109/3DV.2018.00048
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
We present FEATS (Feature Extraction and Tracking Simulator), that synthesizes feature tracks using a camera trajectory and scene geometry (e.g. CAD, multi-view stereo). We introduce 2D feature and matching noise models that can be controlled using a few parameters. We also provide a new dataset of images and ground truth camera pose. We process this data (and a synthetic version) with several current SIM algorithms and show that the synthetic tracks are representative of the real tracks. We then show two practical uses of FEATS: (1) we generate hundreds of trajectories with varying noise and show that COLMAP is more robust to noise than OpenSfM and VisualSIM; and (2) we calculate 3D point error and show that accurate camera pose estimates do not guarantee accurate 3D maps.
引用
收藏
页码:352 / 361
页数:10
相关论文
共 50 条
  • [31] STRUCTURE FROM MOTION
    ULLMAN, S
    JOURNAL OF THE OPTICAL SOCIETY OF AMERICA, 1977, 67 (10) : 1400 - 1400
  • [32] Motion from structure
    Caziot, B.
    Backus, B. T.
    PERCEPTION, 2012, 41 : 79 - 79
  • [33] Extraction and evaluation of kinematic feature values of batting motion from top of swing to ball impact
    Sogo H.
    Henmi T.
    Yoshizawa K.
    Saziki G.
    Iwai Y.
    Tada K.
    IEEJ Transactions on Electronics, Information and Systems, 2021, 141 (03) : 258 - 263
  • [34] Depth Data Assisted Structure-from-Motion Parameter Optimization and Feature Track Correction
    Recker, Shawn
    Gribble, Christiaan
    Shashkov, Mikhail M.
    Yepez, Mario
    Hess-Flores, Mauricio
    Joy, Kenneth I.
    2014 IEEE APPLIED IMAGERY PATTERN RECOGNITION WORKSHOP (AIPR), 2014,
  • [35] Building Better Models: Benchmarking Feature Extraction and Matching for Structure from Motion at Construction Sites
    Cueto Zumaya, Carlos Roberto
    Catalano, Iacopo
    Queralta, Jorge Pena
    REMOTE SENSING, 2024, 16 (16)
  • [36] Articulated motion reconstruction from feature points
    Li, B.
    Meng, Q.
    Holstein, H.
    PATTERN RECOGNITION, 2008, 41 (01) : 418 - 431
  • [37] Learning Structure-from-Motion from Motion
    Pinard, Clement
    Chevalley, Laure
    Manzanera, Antoine
    Filliat, David
    COMPUTER VISION - ECCV 2018 WORKSHOPS, PT III, 2019, 11131 : 363 - 376
  • [38] Multidimensional evaluation of soil pollution from railway tracks
    Małgorzata Wierzbicka
    Olga Bemowska-Kałabun
    Barbara Gworek
    Ecotoxicology, 2015, 24 : 805 - 822
  • [39] Multidimensional evaluation of soil pollution from railway tracks
    Wierzbicka, Malgorzata
    Bemowska-Kalabun, Olga
    Gworek, Barbara
    ECOTOXICOLOGY, 2015, 24 (04) : 805 - 822
  • [40] Simplified Evaluation of Ground Motion Amplification in the Vicinity of a Topographic Feature
    Sierra, Cesar
    Jaramillo, Juan
    Gomez, Juan
    Saenz, Mario
    Vergara, Juan
    JOURNAL OF EARTHQUAKE ENGINEERING, 2022, 26 (13) : 6745 - 6762