DocumentCode
3304038
Title
Determining camera pose using free-form lines
Author
Nagarajan, Sudhagar ; Schenk, Toni ; Csatho, Beata
Author_Institution
Dept. of Geol., Univ. of Buffalo, Buffalo, NY, USA
fYear
2010
fDate
5-5 Nov. 2010
Firstpage
50
Lastpage
53
Abstract
Determining pose or exterior orientation of a camera is an important step in surface reconstruction and subsequent steps such as perceptual organization and object recognition in photogrammetry or computer vision. Pose refers to position and attitude of a perspective camera at the time of exposure. It is generally determined using a set of points in object space and image space. Using free-form lines as a common feature between image and object spaces has always been a challenge. It is basically due to the unavailability of a mathematical model that relates free-form lines in object space, image space and pose of the camera. This paper introduces an innovative mathematical model to determine the camera pose using free-form lines. Everything recorded by a sensor can be considered a discretization of the physical surface. Hence, it is less likely that the same points can be recorded or identified in both image and object space and it is usually easier to find corresponding lines. The method will lead a way to determine pose of images using historical Geographical Information System(GIS) or object space data.
Keywords
cameras; computer vision; feature extraction; geographic information systems; image sensors; mathematical analysis; object recognition; photogrammetry; pose estimation; camera pose determination; computer vision; exterior camera orientation; free-form line; geographical information system data; image sensor; image space; object recognition; object space; photogrammetry; surface reconstruction; Cameras; Equations; Mathematical model; Minimization; Shape; Surface reconstruction; Three dimensional displays; area minimization; exterior orientation; free-form lines; photogrammetry; pose of a camera; registration;
fLanguage
English
Publisher
ieee
Conference_Titel
Image Processing Workshop (WNYIPW), 2010 Western New York
Conference_Location
Rochester, NY
Print_ISBN
978-1-4244-9298-5
Type
conf
DOI
10.1109/WNYIPW.2010.5649747
Filename
5649747
Link To Document