Diploma Thesis
by Denis Klimentjew and Andre Stroh
In the field of autonomous robots, particulary in the domain of humanoid robots,
stereo vision systems are still the most important interface between the robot and
the outside world. The advantage of these systems lies in the possibility to perceive
the environment and to specifically react to events. One of the biggest disadvantages
of this approach lies in the loss of the depth information during image acquisition.
If the depth can be reconstructed, stereo camera systems deliver enough data for
the navigation of robots.
This thesis deals with the theoretical foundations and methods of dimensional recon-
struction, particulary based on the data of a stereo vision system, with the purpose
of determining the depth information. This enables us to assign an unequivocal po-
sition in space to every object in the scene. Based on this, the guiding points will
be detected and their relative position calculated in relation to the robot. Then we
localise the robot on the basis of the available data, i.e., the absolute position of the
guiding points in space.
Hamburg, Germany, January 2008
Downloads
Diploma_thesis.pdf (8.4 MB)
