This week my animation work was put slightly to the side, although
I did continue to work on it, and I worked primarily with projects in the
motion capture lab. At the beginning of
the week we ran a hand capture for a group of researchers. Their hypothesis was that you could
distinguish differences between people by the way in which they type on a
keyboard and once that knowledge was gained, the location of the moving fingers
could be predicted. In order to effectively
capture the data, the cameras in the lab were moved in to create a smaller
capture area. In addition, a different
marker set was used for the hands than I had seen before. There were about 30 smaller markers on each
hand, primarily placed on the joints of the fingers in order to capture the
movement. Another small project I did
some work for this week was for a grad student who works in the graphics
lab. I believe the data that she was
analyzing was for the top portion of a robot that is in the process of being
built at CMU. The robot’s name is
Dorothy and she’s actually a resident of the mocap lab. At the moment she’s only legs, hence the work
being done with upper body movement.
Basically what I had to do was import motion capture data that had been
previously collected and place it onto a character in Maya. There were two different cases that needed to
be juxtaposed against each other- one of a natural running motion and the other
of a constrained run that had been calculated to potentially be used for
Dorothy. At first I ran into difficulties
placing the motion on the skeletons but after doing a little reading about .asf
files I was able to solve the problem.
After that, there was just the simple matter of sub-sampling the natural
motion so that it could be rendered at the same frame rate as the other motion. Then I simply composited
a movie of the two models.