That's what I found as an interim solution.
Each object knows its shape and draws it.
Sensors first find the objects in their surroundings (which is greater than their SensingCoverage) and then ask the object if any part of their area hits the sensor coverage area.
But you know its not a good solution given the Mason classes.
I expect that Continous2D should provide some methods for that reason.
For example when we add an object to Continous2D we should be able to insert its Shape as well.
Thus the methods like getObjectsExactlyWithinDistance would return objects not only by their central coordinations but their Shapes.
This would be a more nice solution.
And I think many Mason users would like it.
Thank you for the suggestion again.
On Tue, Dec 21, 2010 at 7:41 PM, Mike Little <[log in to unmask]>
Here's an alternative thought that you might find simpler than
JTS - why not store the area of the object (car, bus) as a property
of the object and have the sensor query this property and "respond"
> On Tue, Dec 21, 2010 at 4:11 PM, Mark Coletti <[log in to unmask] <mailto:[log in to unmask]>>
On 12/21/2010 5:38 PM, murat karakaya wrote:
> Thank you for the reply,
> however, I'm looking for a simpler solution -in terms of learning time of new classes-
> <mailto:[log in to unmask]
> > I want to simulate a sensor network composing of static sensor nodes which
> > monitor a road for passing-by objects. These mobile objects are in different
> > shapes and areas. I want to sensor nodes would sense the object not by its
> > center coordinates but by its area. For example a car object would have a
> > small rectangle area wrt that of a bus object.
> You may want to look into GeoMason. Though the intent with GeoMason
> is to deal with geospatial entities, it is wrapped round the Java
> Topology Suit (JTS). JTS provides the type of functionality you are
> looking for, including querying Geometry area.
> GeoMason is found in the "contrib" directory in the main MASON distribution.