Binocular Computer Vision Based on Conformal Geometric Algebra

被引：16

作者：

Hrdina, Jaroslav ^{[1
]}

Navrat, Ales ^{[1
]}

机构：

[1] Brno Univ Technol, Fac Mech Engn, Inst Math, Brno, Czech Republic

来源：

ADVANCES IN APPLIED CLIFFORD ALGEBRAS | 2017年 / 27卷 / 03期

关键词：

Conformal geometric algebra; Clifford algebra; Binocular vision; Projective geometry;

D O I：

10.1007/s00006-017-0764-4

中图分类号：

O29 [应用数学];

学科分类号：

070104 ;

摘要：

We apply the conformal geometric algebra (CGA) to the generalized binocular vision problem. More precisely, we reconstruct a 3D line from its images on the image planes of two cameras whose mutual position is specified by a given Euclidean transformation which depends on an arbitrary number of parameters. We represent all transformations by CGA elements which allows us to derive the general equations of 3D line reconstruction by formal CGA elements manipulation. The transformation equations can be solved w.r.t. either motor or projection unknown parameters. We present two specific examples, show the explicit form of two particular motors and solve the appropriate equations completely.

引用

页码：1945 / 1959

页数：15

共 8 条

[1]

Ablamowicz R., 2015, CLIFFORD BIGEBRA MAP

[2]

Dorst Leo, 2009, Geometric Algebra for Computer Science (Revised Edition): An Object-Oriented Approach to Geometry

[3] Robust Pose Control of Robot Manipulators Using Conformal Geometric Algebra [J].