An iterative method for 3D reconstruction under orthography

被引：0

作者：

Kaliva, E ^{[1
]}

Delopoulos, A ^{[1
]}

机构：

[1] Aristotle Univ Thessaloniki, Dept Elect & Comp Engn, GR-54006 Thessaloniki, Greece

来源：

PROCEEDINGS OF THE FIFTH IASTED INTERNATIONAL CONFERENCE ON VISUALIZATION, IMAGING, AND IMAGE PROCESSING | 2005年

关键词：

3D motion; 3D structure; 3D reconstruction; structure from motion; orthography; geometric algebra;

D O I：

暂无

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Three-dimensional object reconstruction and the corresponding estimation of motion and shape from two-dimensional projections is a task of great importance in the fields of digital video processing, computer vision and specific areas of biomedicine, robotics, topography etc. Present work presents an iterative method for estimating the three-dimensional parameters needed for reconstructing the observed object. The geometric objects and their relationships have been modelled as elements of a 3D Geometric Algebra. Extensive experiments indicate that the proposed algorithm is more robust to measurement noise - such as motion estimation errors - than traditional tools for extracting "structure from motion" (SFM) under the orthography assumption.

引用

页码：382 / 387

页数：6

共 17 条

[1] Geometric algebra: A computational framework for geometrical applications - Part 1 [J].

Dorst, L ;

Mann, S .

IEEE COMPUTER GRAPHICS AND APPLICATIONS, 2002, 22 (03) :24-31

[2]

Hestenes D., 2012, CLIFFORD ALGEBRA GEO

[3] MOTION AND STRUCTURE FROM FEATURE CORRESPONDENCES - A REVIEW [J].

HUANG, TS ;

NETRAVALI, AN .

PROCEEDINGS OF THE IEEE, 1994, 82 (02) :252-268

[4] MOTION AND STRUCTURE FROM ORTHOGRAPHIC PROJECTIONS [J].

HUANG, TS ;

LEE, CH .

IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 1989, 11 (05) :536-541

[5]

HUANG TS, 1986, HDB PATTERN RECOGNIT, pCH14

[6] New geometric methods for computer vision: An application to structure and motion estimation [J].

Lasenby, J ;

Fitzgerald, WJ ;

Lasenby, AN ;

Doran, CJL .

INTERNATIONAL JOURNAL OF COMPUTER VISION, 1998, 26 (03) :191-213

[7]

MA Y, 2004, INVITATION 3D VISION

[8] Geometric algebra: A computational framework for geometrical applications Part 2 [J].

Mann, S ;

Dorst, L .

IEEE COMPUTER GRAPHICS AND APPLICATIONS, 2002, 22 (04) :58-67

[9] Motion from three weak perspective images using image rotation [J].

Ostuni, J ;

Dunn, S .

IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 1996, 18 (01) :64-69

[10] A paraperspective factorization method for shape and motion recovery [J].

Poelman, CJ ;

Kanade, T .

IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 1997, 19 (03) :206-218

← 1 2 →