In this test I used the depth map from jit.freenect.grab to create an OpenGL mesh. "Shadows" or areas where the Kinect wasn't able to compute depth create some jarring artifacts but I decided to try to use them to my advantage aesthetically.
@mattovision Actually, that won't help. The problem is that the Kinect projects a pattern of infrared red light that it uses to estimate depth. Wherever the light doesn't reflect, or is blocked by something, it fails to measure the depth. What you're seeing is in fact me playing with my own shadow, which has now become a solid object thanks to OpenGL.
@mattovision Actually, that won't help. The problem is that the Kinect projects a pattern of infrared red light that it uses to estimate depth. Wherever the light doesn't reflect, or is blocked by something, it fails to measure the depth. What you're seeing is in fact me playing with my own shadow, which has now become a solid object thanks to OpenGL.
jovansystem 1 year ago
@jovansystem
Ah in that case you need to steal the lighting rig from a glamour photographer.
mattovision 1 year ago
@mattovision No, the values are so far off the chart that low-pass filtering doesn't cut it.
jovansystem 1 year ago
Quite unique and beautiful.
I think some low-pass filtering would remove the noise, if you wanted to remove it that is.
mattovision 1 year ago