Robot navigation landmarks
-
usage of the affine moment invariants
(AMIs) for robot navigation
-
landmark based navigation. Patterns
formed by two concentric circles with equal outer and different inner radii
are chosen as a mark set (see Fig. below). The information about its position
is encoded into the ratio of its inner and outer radii. The dimension indirect
use for the coding ensures the independence of the code on the robot distance
from the mark. The mark shape is simple enough and easy to detect.
-
AMIs - moment-based features for description
and recognition of geometrically distorted objects and patterns, based
on central moments of the images function
-
stability of the AMIs with respect
to perspective projection of the image as well as to additive random noise
the computer-simulated robot vision
system and artificially generated mark distortions were used to test the
stability of the AMIs under perspective projection of the marks and noise
corruption.
Experiment arrangement
-
real data taken by the camera were
used to demonstrate the AMIs' ability to make a correct classification
of distorted marks in a real environment. The images of the marks with
the same inner radii but taken with the different viewing angles have their
AMI values very close in comparison to the AMI values distances of
the images, when the marks have
different inner radius.
Relevant publications: