il 2004
A UNIFIED FRAMEWORK FOR THE AUTOMATIC MATCHING OF POINTS AND LINES IN
MULTIPLE ORIENTED IMAGES
Christian Beder
Institute of Photogrammetry, University of Bonn
Nussallee 15, 53115 Bonn, Germany
beder@ipb.uni-bonn.de
KEY WORDS: Matching, Reconstruction, Statistics, Geometry, Algorithms
ABSTRACT
The accurate reconstruction of the three-dimensional structure from multiple images is still a challenging problem, so
that most current approaches are based on semi-automatic procedures. Therefore the introduction of accurate and reliable
automation for this classical problem is one of the key goals of photogrammetric research.
This work deals with the problem of matching points and lines across multiple views, in order to gain a highly accurate
reconstruction of the depicted object in three-dimensional space. In order to achieve this goal, a novel framework is
introduced, that draws a sharp boundary between feature extraction, feature matching based on geometric constraints and
feature matching based on radiometric constraints. The isolation of this three parts allows direct control and therefore
better understanding of the different kinds of influences on the results.
Most image feature matching approaches heavily depend on the radiometric properties of the features and only incorporate
geometry information to improve performance and stability. The extracted radiometric descriptors of the features often
assume a local planar or smooth object, which is by definition neither present at object corners nor edges. Therefore it
would be desirable to use only descriptors that are rigorously founded for the given object model. Unfortunately the task
of feature matching based on radiometric properties becomes extremely difficult for this much weaker descriptors.
Hence a key feature of the presented framework is the consistent and rigorous use of statistical properties of the extracted
geometric entities in the matching process, allowing a unified algorithm for matching points and lines in multiple views
using solely the geometric properties of the extracted features. The results are stabilized by the use of many images
to compensate for the lack of radiometric information. Radiometric descriptors may be consistently included into the
framework for stabilization as well.
Results from the application of the presented framework to the task of fully automatic reconstruction of points and lines
from multiple images are shown.
1 INTRODUCTION
This work deals with the matching of points and lines across
multiple images. This problem has been addressed exten-
sively for points (c.f. (Schmid and Mohr, 1997) or the
classical textbooks (Horn, 1986) and (Faugeras, 1993)) and
also for lines in (Schmid and Zisserman, 1997), (Baillard
et al., 1999) and (Heuel and Forstner, 2001). All of those
approaches generate matching pairs of features from the
image intensity data and then use the known geometric in-
formation for forward intersection in order to obtain a 3D
reconstruction of the depicted object. If you consider a
situation like in figure 1, where a simple cube is depicted
from all its six sides, it is obvious, that all those feature
matching methods relying on radiometric information in
the first place must fail, since no face ofthe object is visible
in more than one view. On the other hand, all line features
are visible in two views and all point features are visible
in three views, thus a precise 3D reconstruction should be — Figure |: A cube depicted from all its six sides. No pair
possible given the matches. The problem is of course the = of pictures shows the same face, though a reconstruction is
assumption, that the observed surface is local planar at the possible given the orientation of the cameras.
features, which is by definition neither the case at object
corners nor at object edges, that are of primary interest for
an accurate scene reconstruction. Using the known image
geometry has been applied by (Jung and Paparoditis, 2003)
for edgel matching across multiple views.
1109
