Detailed real-time urban 3D reconstruction from video

被引：426

作者：

Pollefeys, M. ^{[1
]}

Nister, D. ^{[1
,2
]}

Frahm, J. -M. ^{[1
]}

Akbarzadeh, A. ^{[2
]}

Mordohai, P.

Clipp, B. ^{[1
]}

Engels, C. ^{[2
]}

Gallup, D. ^{[1
]}

Kim, S. -J. ^{[1
]}

Merrell, P. ^{[1
]}

Salmi, C. ^{[1
]}

Sinha, S. ^{[1
]}

Talton, B. ^{[1
]}

Wang, L. ^{[2
]}

Yang, Q. ^{[2
]}

Stewenius, H. ^{[2
]}

Yang, R. ^{[2
]}

Welch, G. ^{[1
]}

Towles, H. ^{[1
]}

机构：

[1] Univ N Carolina, Dept Comp Sci, Chapel Hill, NC 27515 USA

[2] Univ Kentucky, Ctr Visualizat & Virtual Environm, Lexington, KY 40536 USA

来源：

INTERNATIONAL JOURNAL OF COMPUTER VISION | 2008年 / 78卷 / 2-3期

关键词：

3D reconstruction; stereo vision; structure from motion; large scale modeling; urban reconstruction; plane sweeping; depth map fusion;

D O I：

10.1007/s11263-007-0086-4

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 presents a system for automatic, geo-registered, real-time 3D reconstruction from video of urban scenes. The system collects video streams, as well as GPS and inertia measurements in order to place the reconstructed models in geo-registered coordinates. It is designed using current state of the art real-time modules for all processing steps. It employs commodity graphics hardware and standard CPU's to achieve real-time performance. We present the main considerations in designing the system and the steps of the processing pipeline. Our system extends existing algorithms to meet the robustness and variability necessary to operate out of the lab. To account for the large dynamic range of outdoor videos the processing pipeline estimates global camera gain changes in the feature tracking stage and efficiently compensates for these in stereo estimation without impacting the real-time performance. The required accuracy for many applications is achieved with a two-step stereo reconstruction process exploiting the redundancy across frames. We show results on real video sequences comprising hundreds of thousands of frames.

引用

页码：143 / 167

页数：25

共 69 条

[1]

AKBARZADEH A, 2006, P INT S 3D DAT PROC

[2]

*AM SOC PHOT, 2004, MAN PHOT

[3]

[Anonymous], EUR C COMP VIS

[4]

[Anonymous], IEEE C COMP VIS PATT

[5]

[Anonymous], MACHINE VISION APPL

[6]

[Anonymous], 1993, Three-Dimensional Computer Vision: A Geometric Viewpoint

[7] RECURSIVE ESTIMATION OF MOTION, STRUCTURE, AND FOCAL LENGTH [J].

AZARBAYEJANI, A ;

PENTLAND, AP .

IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 1995, 17 (06) :562-575

[8]

Baker S, 2003, CMURITR0301

[9] Sequential updating of projective and affine structure from motion [J].

Beardsley, PA ;

Zisserman, A ;

Murray, DW .

INTERNATIONAL JOURNAL OF COMPUTER VISION, 1997, 23 (03) :235-259

[10]

BIBER P, 2005, ISPRS WORK GROUP V4

← 1 2 3 4 5 6 7 →