|
|
|
|
|
|
|
chris...
Joined: 20 Oct 2010
Posts: 28
|
My opinion, mostly to do with the difference between something like robosoccer and robowars is that robowars is war. With no markers it will be more complex to initially implement but they wont get damaged, break off or need to be added
I can see a few problems with kinect, its extremely laggy, has a limited operting range, a narrow field of view and costs a bit.
Personally I wouldn't bother with markers or kinect, but that's my opinion, just remember what ever gets implemented would most likely become a standard or a base and mostly likely be used by most.
Using ssl vision all robots will need a constant unobstructed view of the markers.
ROS is interesting contains a lot of re-usable code, no point in writing something again that already been written.
Their are two solutions for tracking
1. Something like ssl vision where all robots know the absolute position of themselves and all other robots, simply solution
2. Each robot works out the angle and distance of the other robots
The second option will be better and allow self contained automation
http://robotbox.net/blog/gallamine/open-lidar-project-hack-neato-xv-11-lidar-200-bounty
Last edited by chris... on Tue Nov 30, 2010 9:02 am; edited 1 time in total
|
Tue Nov 30, 2010 8:30 am |
|
|
|
|
|
|
|
|
|
|