Page 176 - Introduction to Autonomous Mobile Robots

161
Perception
x
The selection of all line segments that contribute to the same line can now be done in
j
a threshold-based manner according to
T
( x – x) x – x) ≤ d (4.80)
(
j j m
where d m is a threshold value and is the representation of the reference line (from a
x
model, average of a group of lines, etc.).
But the approach of equation (4.80) does not take into account the fact that for each mea-
surement and therefore for each line segment we have a measure of uncertainty. One can
improve upon this equation by selecting line segments that are weighted by their covariance
matrix C :
j

–
1
T
( x – x) C + C) ( x – x) ≤ d m (4.81)
(
j
j
j
The distance measure of equation (4.81) discriminates the distance of uncertain points
in model space considerably more effectively by taking uncertainty into account explicitly.
4.3.1.3 Range histogram features
A histogram is a simple way to combine characteristic elements of an image. An angle his-
togram, as presented in figure 4.39, plots the statistics of lines extracted by two adjacent
range measurements. First, a 360-degree scan of the room is taken with the range scanner,
and the resulting “hits” are recorded in a map. Then the algorithm measures the relative
angle between any two adjacent hits (see figure 4.39b). After compensating for noise in the
readings (caused by the inaccuracies in position between adjacent hits), the angle histogram
shown in figure 4.39c can be built. The uniform direction of the main walls are clearly vis-
ible as peaks in the angle histogram. Detection of peaks yields only two main peaks: one
for each pair of parallel walls. This algorithm is very robust with regard to openings in the
walls, such as doors and windows, or even cabinets lining the walls.

4.3.1.4 Extracting other geometric features
Line features are of particular value for mobile robots operating in man-made environ-
ments, where, for example, building walls and hallway walls are usually straight. In gen-
eral, a mobile robot makes use of multiple features simultaneously, comprising a feature
set that is most appropriate for its operating environment. For indoor mobile robots, the line
feature is certainly a member of the optimal feature set.
In addition, other geometric kernels consistently appear throughout the indoor man-
made environment. Corner features are defined as a point feature with an orientation. Step
discontinuities, defined as a step change perpendicular to the direction of hallway travel,