Hello,

Right now I have a camera on a robot running ROS, and I'm trying to
figure out how to set things up to meet my needs. 

1) The robot currently has a USB camera but may need to use Firewire in
the future. Ideally whatever solution should work for both cases. The
baseline requirement is to get a 320x240x10fps color video or image
stream.

2) I have 1-2 vision tasks I would like to be performed locally on the
robot.

3) I would like to transmit the video/image stream to a ground station
using TCP/IP so I don't have multiple radios on the robot.

4) The ground station should be able to display the video/images and
perform computer vision tasks on the data.

5) The video should be logged for disk so that it can be fed back into
the system for debugging.

It looks like there are a few different ways to solve my problems but I
am wondering if there is a suggested method that will allow for the
maximum code reuse.

Right now I am using gstreamer but it does not seem like it will scale
up cleanly to multiple vision tasks and the configuration system for
gsreamer is less than ideal. Should I be using image_pipeline, nodelets
or something else?

It looks like the Texai are using Skype for tele-operation, is there a
planned migration to a unified image/video system?

Thanks,

Bill