... the inter -robot formation.
Distributing the group control tasks to individual robots must be
compatible with the control and sensing capabilities of the individual
robots. As the inter -robot ... with
inter -robot constraints, determine whether there exist robot trajectories
that maintain those constrains. For both directed and undirected (not
necessarily rigid) formations...
... a single robot.
Individual robots typically obtain partial and noisy data from the
surrounding environment. This data is often erroneous, leading to
miscalculations and wrong behaviours, and ... the N robots by the
M sites) to a final or target configuration.
x A formation feasibility problem: given the kinematics of several
robots along with inter -robot constraints, determine...
... time. In general these point-based landmarks are corners, van-
ishing points and points with distinct colour and contrast in the environ-
ments.
Unlike point-based landmarks, scene-based landmarks ... sensors [32 ]. Alternatively combining vision
sensor and dead reckoning [33 ], [29].
2 .3 Real-time Visual 2-D Landmark Recognitions
This section describes the development of a 2-D...
... graphical en-
hancement of visual cues and out-of-view data. This will assist the opera-
tor in interpretation and command issue thus increasing productivity and
reducing fatigue.
The system concept ... distances, yet are relatively insensi-
tive to translation. In fact, MRF relates the spacial and temporal informa-
tion together, to find the most likely displacement between im...
... feature matching algo-
rithm was worst at finding the best match between an image in the training
Using Visual Features for Building and Localizing within Topological Maps 2 63
7.4.2 Determining the ... set 2 were taken on a zigzag path across the training path.
5
1
2
3
4
6
7
8
9
10
11
12
Start
13
14
15
16
17
18
19
20
21
22
23
24
25
T 2-0
T 2-1
T 2-2
T2 -3
T2 -4
T2...
... essentially two major types of maps; geomet-
rical and topological maps. Geometrical maps provide details of metrical
information (exact co-ordinates and distances) between objects found in
the ... the 3- D external environment. As a result, depth
Q.V. Do et al. : Vision-Based Autonomous Robot Navigation, Studies in Computational Intelli-
www.springerlink.com
c
Springer-Ver...
... on Intelligent Robots and Sys-
tems, pp.88 3- 8 88, 2001.
19. D. Jung and A. Zelinsky, "Integrating spatial and topological navigation in a
behaviour-based multi -robot application," in ... self-organizing
neural network architecture for fast supervised learning and pattern recogni-
tion," in Proc. International Joint Conference on Neural Networks, IJCNN-9 1...
... vector is normalised so the in-
puts lie in the range [0,1].
Table 5 .3 Typical rule-base and rule confidences obtained after training. Rule confidences are shown in brackets in the
following order: ... experimental conditions were studied:
i. off-line supervised learning with back-propagation training;
ii. on-line reinforcement learning;
iii. hybrid of supervised and reinfo...
...
129–1 43.
52 D. Van Zwynsvoorde, T. Simeon, and R. Alami, Incremental topo-
logical modeling using local voronoi-like graphs, EEE/RSJ Interna-
tional Conference on Intelligent Robots and Systems, ... Modelling 233
Supposedly the robot is modelled as a point operating in a subset belong-
ing to the two-dimensional Euclidean space (in our particular case, al-
though this is true w...