This project is a collection of tools and demos for controlling a Baxter robot.

conv_tools contains functions for converting between various position and
message formats. It is primarily meant to be used by pos_tools

pos_tools contains functions for moving Baxter and retrieving position 
information. In progress is a concept to incorporate force interrupts
(i.e. having Baxter stop moving when a certain force level is measured)
by running movement and force monitoring in separate threads. Had problems
trying to initialize rospy twice (in order to start nodes in both threads),
but it should clearly be possible to get around this - the point of ROS is
having multiple nodes running.

vis_tools contains functions for image processing. It can edit a gripper out
of an image, find dark object in a black and white image, recognize a specific
object in a scene. Uses OpenCV 3. Uses the Python multiprocessing library in 
the image recognition code to avoid crashing the entire program when the C++
code in OpenCV segfaults. (Would be preferable to not have segfaults, but this 
works.)

fetch_cp is a demonstration where Baxter locates a specific object, picks it up
and brings it back to a preset location. In testing, a white box was used with
a design drawn in Sharpie.

square_finder is a demonstration where Baxter moves a piece of glass from a red
square to a blue square using gecko grippers. In testing, the background was 
plain white and the squares were cardboard with white infill and colored tape 
around the borders. Missing the lowest-level pickup and putdown functions for 
the gecko grippers, but they were specific to the mechanical design, so no 
great loss. Of interest in this file is more the computer vision.

The research was carried out, in part, at the Jet Propulsion Laboratory, 
California Institute of Technology, under a contract with the 
National Aeronautics and Space Administration. Copyright 2015 
California Institute of Technology. Government sponsorship acknowledged.