ul 2004
ats and
lights).
| at the
ix, and
d down
'ntation
ee and
itil the
several
epeated
events
by the
| or by
several
these
nd and
g-point
with a
d mesh
enerate
eck the
€
gn the
viously
uitable
nent a
metric
hrough
natural
nted a
15, and
ate the
, or to
three
ve, and
;
bserve
n also
à and
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B2. Istanbul 2004
Boats may have different directions or trajectories, and speeds.
They can be added, selected and deleted, or configured by the
user. Animation can start or stopped at any time. All boats may
move simultaneously, and then the model becomes something
like a live map, showing real-time changes. The model may still
be manipulated in the same way, as when animation is off. Each
boat can be controlled and navigated by the user, who can
change its speed and direction, stop it or set its destination
point.
There are four different camera modes. First is the general view
(Figure 1). Other modes are boat-dependent, and to use them
one of the boats has to be selected first. The camera is then set
in a position dependent on the selected boat's location, and
moves with the boat through the scene. It can be placed behind-
and-above the boat, behind the boat under the water, or on
board. In these modes the user can look around and control the
boat to turn left or right, to speed up or slow down, to stop or to
move. Figure 2 shows the view from behind and above the boat.
Fidel Satie
Position (WGSS4): 22°14,009 N 114°7.1054 E 207
Figure 2: MGIS Model — Scene of Navigational Mode.
When animation is activated, the viewpoint will follow
the movement of the ship model. The movement of the
vessel could be controlled by using mouse clicks.
The system can also display the Voronoi mesh on the sea
surface. In this mode points inside the boats, points on the
coastline, and Voronoi diagram edges can be seen, and the user
can watch real-time Voronoi diagram changes, as points move
through water mesh (Figure 3). This mesh is used for collision
detection and obstacle avoidance.
Figure 3. The Marine GIS in Voronoi diagram display mode
691
6. IMPLEMENTED FEATURES
e Depth testing
Each boat can monitor the depth of water under itself. The
depth can also be checked at any point on the water surface.
Given the desired location, the height of the underlying mesh at
that point can be obtained by using natural neighbour
interpolation (Sibson, 1981, Gold, 1989).
e Deepest channel navigation
Because of the depth monitoring feature, and the automatic boat
control, the boat can be navigated to the deepest channel over
the chosen path between the current location and a given target
point.
e Collision detecting and preventing
While moving a boat through the Voronoi mesh, collision with
another point could occur. The mesh can detect this and initiate
steps to avoid it. It could be enough merely to change the
point’s/boat’s direction, or speed. If a collision with another
boat is likely then the directions and speeds of both boats can be
changed, and restored after boats safely pass each other. If a
coastline data point is an obstacle the boat should be redirected
to prevent running aground.
e Changing water level (handling of 2 meshes)
Changing the water level, to simulate tides, can be done even
during the animation. The water mesh has to be reconstructed
after such a change, as the coastline points will change. The
underlying mesh has to be scanned, and every triangle of it has
to be checked to see if it is being intersected by the horizontal
plane of the new water level. It is necessary for points to be
distributed quite uniformly and frequently enough along all
coastlines, to make sure that no boat can pass through the
coastline without a collision being detected.
7. A PILOT APPLICATION — THE “PILOT BOOK”
Perhaps the ultimate example of a graphics-free description of a
simulated real world is the Pilot Book, prepared according to
international hydrographical standards to define navigation
procedures for manoeuvring in major ports (UK Hydrographic
Office, 2001). It is entirely text-based, and includes descriptions
of shipping channels, anchorages, obstacles, buoys, lighthouses
and other aids to navigation. While a local pilot would be
familiar with much of it, a foreign navigator would have to
study it carefully before arrival. In many places the approach
would vary depending upon the state of the tides and currents. It
was suggested that a 3D visualization would be an advantage in
planning the entry into the harbour. While it might be possible
to add some features to existing software, it appeared more
appropriate to develop our own 3D framework. (Ford, 2002)
demonstrated the idea of 3D navigational charts. This was a
hybrid of different geo-data sources such as satellite pictures,
paper chart capture and triangular irregular network data
visualized in 3D. The project concluded that 3D visualization of
chart data had the potential to be an information decision
support tool for reducing vessel navigational risks. Our
intention was to adapt IHO S-57 Standard Electronic
Navigation Charts (International Hydrographic Bureau, 2000,
