A Robust Feature Matching Method for Robot Localization in a Dynamic Indoor Environment - Technologies and Applications of Artificial Intelligence - page 354

An example map of a gi
are the possible location of

ven environment is shown in Fig. 1. The hypothesis po
the robots as shown in Fig. 2.

oints

Fig. 1. Given ma ap Fig. 2. Hypothesis points

2.2 Real-Time Feature Generation

The robot with laser range
Fig. 3. We will resample t
vector, as shown in Fig. 4
feature points (HFP) inform

efinder can generate the features at real-time as shown
the information and get a reasonable size of informat
4. Global hypothetical features are the 360 ° hypothet
mation a robot might get in a given environment.

n in
tion
tical

b

a

Fig. 3. a. Robot in an indoor e envirionment. b. The information gathered by a laser rangefind der.

b

a

Fig. 4. a. the environment ac ccording to the information in Fig 3. b. local features of a robo ot.

2.3 Omni-directional F Feature Grouping

In this paper, we divide the
groups, and repeat the fe
hypothetical feature point (
contains 180 ° length infor

e information gathered by the laser rangefinder into seve
ature matching for each group. We divide each of
(HFP) into 8 groups as shown in Fig. 5. Each of the gr
rmation gathered by the laser rangefinder.

eral
the
oup