|
Class : Habitué
Posts : 66
Registered on : 28/08/2007
Off line
|
Following on from the undoubted success of my fibre optic thread, I present to you, the sequel..
Is there anything to be gained by having two webcams? Currently, my single VX1000 sits on top of my monitor, directly in front of me. How about if we could have two webcams? If I were to put them on top of my monitor, but at each end, they'd be 50cm apart. Obviously, you'd angle them inwards slightly.
So, if Freetrack works by working out where the emitters are in space, surely having two would make this more accurate and potentially, easier to work out?
Would there be a CPU overhead?
Is the cost prohibitive?
Would it allow for a model dimensions self calibration mode?
Clearly the software doesn't currently support this. I wonder how much of a rewrite would be involved?
|
|
Class : Webmaster (admin)
Posts : 780
Registered on : 13/07/2007
Off line
|
FreeTrack uses weak perspective algorithms, so cannot correlate data between multiple cameras. Full perspective algorithms can do this but are more computationally intensive, use less reliable iterative methods that aren't guaranteed to converge and depend on knowledge of the intrinsic camera properties (focal length, sensor size and lens distortion).
|