I spent all day trying to solve a problem, so maybe someone have any ideas of how to help me.
I use Kinect node Player for detecting a person, and Edge.EX9 + Threshold.EX9 for detecting an edges of player. Then, i use a pipet for detecting coordinates of white outlines, but here comes the problem. All i want in a result is a finite numbers of coordinates, that i can use for some lines further.
The problems is that this number of coordinates is not constant because of noise, and when i use them for lines (which must be drawn from current position to position in previous frame) some of them lost and lines draws all over the screen like crazy.
SmoothSettings on Kinect node and some Blur didn’t help. Is there any ways to solve that problem, without writing my own blob detection algorithm (as in this nice article http://www.creativeapplications.net/processing/kinect-physics-tutorial-for-processing/ )?
I would be very grateful for any clue in which way i have to think.