This page is a record of the experimental run of the vision scout grabbing images as it drives approximately 15 feet across the floor.
Experiment of September 6, 2002.
This experiment was arranged in SBP lab 2 here at Carnegie Mellon.
The purpose of the experiment was to test if we can grab frames from the firewire camera/ omnidirectional mirror setup as the scout drives across the room.
Landmarks of flourescent pink boxes were placed in the room to demonstrate what the camera will try to pick out in order to localize.
The pictures were saved in portable pixel map format allowing for ease of reading the files into another program to process them.
The pictures below were saved in rgb ppm format and then converted to jpg for the web.
The experiment was run again in order to grab a second set of images, this time in yuv ppm format for post processing.
Next, these images will be post processed to determine landmark locations in the image plane. They will be fed, off-line into the slam algorithm in order to test and debug the system.
The following 6 images show samples of the set of 36 images as the robot traversed its path. The direction of the robot's motion is towards the top of the pictures. The landmark boxes can be easily distinguished.
|
|
|
|
|
|
Go back to: H-SLAM Experiments Page | H-SLAM page
Jump to: People | Robots | Research | Events | Projects | Education | Papers | Software | Links
© Copyright 2002 Sensor Based Planning Lab, Carnegie Mellon University. All Rights Reserved.
 |
 |
 |