 |
|
Alley Experiment [MPEG movie (5.5 MB)]
|
In this experiment, the target performs two trajectory that are identical (up to small approximation). It first moves through a long alley made up with cardboard panels and then makes a right turn around a square column. For each trajectory, the observer tracks the target with a different strategy -- first pure visual servoing and then our track algorithm. (Watch the video).
This experiment consists of three parts. In each part, the target is guided through a similar trajectory with a permanent obstacle (a square cardboard box) placed in the middle, and the observer follows. Each part differs from the other two in the way we place the second obstacle (another box) --- in Part (a), we do not place the second box; in Part (b), we place the second box at halfway through the tracking process and leave it there; and in Part (c), we leave the same second box only for a short time (notice a short swerve of robot's heading) . This experiment is aimed at showing the tracking system's ability to adjust to changes in the environment.
In this experiment, the observer follows the target performing a long tour starting outside one of the offices along the south corridor of the laboratory. The observer tracks the target down the south corridor, through a lounge area cluttered with chairs, and into the north corridor. (Watch the video)
Thursday, January 09, 2003 02:46:06 PM
