Hey guys, long time no see. Guess that’s a sign that I’m improvvvving in the art of patching. :)
What I want is to put an IR camera above a area in front of a big LED panel to track the X and Z positions of a person (i.e., if the person is going to towards the panel or backwards, and left/right). So far so good, but now I have to track the height of the person too. They have to reach the hand above his head to activate some commands on the LED panel. So, my idea was to put another camera in the back of this area, facing the giant LED panel, to track them.
Then comes my problem. Is it possible to use the outputs of the two Contour nodes and link them together? I know I can link them using the X coordinate of both cameras, but I don’t think that is reliable enough. Another problem is when a person pass in front of the other in the second camera’s point of view, which can mess up the linking, but in my application that is not a problem at all.
I’m asking for ideas on how to do this, I know it’s not a particular patching question. :)
Another interesting way should be stereoscopy. The goals should be to calculate a disparity map (a grayscale image that rapresent depth of the objects in the scene). Than you can apply a threshold and some blob tracking teqniques to determine the z of obecjt or to recontruct in 3d the stereo scene.
For doing that, some time ago i’ve took a look to sentience, a library in c# for robotic vision.
but i am too uncultured for programming that things :P
i hope someone of you, incredible developers, is interested in this way, to develop a module of stereoimaging in vvvv ;)