Feature-based Photogrammetr ic Object Reconstruction (I)
Date Issued
2002
Date
2002
Author(s)
DOI
902211E002067
Abstract
The point feature has been inherently
utilized for photogrammetric tasks,
and topographic mapping, etc. Due to the
under expectation. One can easily
meaningful point feature from a noise by observing a distinct image point. From
computer vision point of view, the point
feature is of low-level. The degree of
automation using only point features
remains unsatisfactory, leading to heavy
human’s intervention in mapping tasks.
On the other hand, from the object
reconstruction view point, the linear feature
that can be easily extracted from the image
via image processing techniques is of
higher-level information than its counterpart,
the point feature. It is especially
advantageous of employing linear features
when considering the urban photographs in
which the man-made buildings,
polyhedron-type, are the main structure in
the scene. The fitted line by adjusting many
points which are regarded as if they are
along the line would reasonably suppress
the noise effect and provide linear feature
with higher precision. This study was
motivated by the above facts and aimed at
solving for photogrammetric orientation and
performing three-dimensional object
reconstruction by coplanar linear features.
An experimental test on an aerial stereopair
over urban area has been conducted for the
purposes of solving orientation and
demonstrating object reconstruction.
Subjects
Object reconstruction
Linear
feature
feature
Coplanar linear features
Publisher
臺北市：國立臺灣大學土木工程學系暨研究所
Type
report
File(s)![Thumbnail Image]()
Loading...
Name
902211E002067.pdf
Size
468.13 KB
Format
Adobe PDF
Checksum
(MD5):7483f8acf73d992d6e9ccb3bc144b715