|
[DPRG] Re: Games
Subject: [DPRG] Re: Games
From: David M Wilson
davidmw at tx.rr.com
Date: Tue Jun 19 18:58:02 CDT 2007
>>Here are some images from The Science Place of the floor
in the lobby. I thought they might be useful for visualizing
how the space can be used as a real-world robot environment.
What a neat floor - and the CdS cells on the claw sure worked well.
I'm just chiming in with some feedback about Drifter's chances of detecting
the lines on that floor with the camera.
The current vision system is exclusively blob based and it expects well
defined and saturated hues on the target. Pushing the system to detect off
white and white, or black, involves shifting the detectable color model on
the brightness axis to such an extent that brightness info overwhelms hue.
Every shadow then looks like a black target. At the other end of the scale
every reflection, light source, white wall, and light color will merge to
become a target blob.
Using the camera Drifter would follow the those lines for a short distance
and then would take off chasing a reflected light source. Shiny black looks
nearly identical to white and this floor will likely need a different vision
library or some line sensors.
David Wilson
More information about the DPRG mailing list
|
