Over Exposure

[purduephotog]

Originally Posted by ChaosX73

Hi! I am a member of team 2338. Our alliance won 2nd place at the Chicago Regionals. We had problems with our camera and we believe that this is due to over exposure. Because of the black curtains and large spotlights, when we ran the calibration for our camera, all that came up were small, very thin rectangles of coorsponding colors (please excuse my spelling). We have some ideas on how to fix this problem, but a little advice will help a lot. Can anyone help us?

Does the 206 not allow you to set pre-programmed gain/offset information on it? (ie turn off auto-exposure).

[Greg McKaskle]

Since the 206 is a security camera, it has lots of auto settings, and has some ability for static setup, but typically no numerical values. For exposure you choose between auto and hold. This would technically allow you to set and hold the exposure, but you have no way to directly set or read the actual exposure time. You'd need a calibration procedure to "set" the exposure. I've done this for quick experiments, but it is not very repeatable. The gain control is referred to as exposure priority, and the choices are quality and speed.

Since the lighting changes so much as you pivot on the field, I don't believe a fixed exposure will work well.

See videos such as http://www.youtube.com/watch?v=xJnq-...eature=channel

for a robots eye view of the field. This was not taken with the Axis 206, but with some sort of hand held video cam.

As for the polarized sunglass idea, again, good idea, but that was what I was referring to. If the reflected light isn't polarized, they a polarizer won't seem to affect the glare any differently than the rest of the scene. It will simply darken the scene evenly by some amount. If in auto, the camera will then adjust the exposure and you are back to where you started. Barn doors can limit the direct view of the lights, but unfortunately, they cannot help with the reflected light from the targets.

I'm not trying to discourage investigation here, just giving results on things I've been trying. If anyone else has ideas I'm game. I'll gladly test it out and try to make it effective.

Greg McKaskle