Taggart Bonham, a Computer Science Major at Dartmouth College, has worked for several years in Corporate R&D at Magna International focusing on the application of computer vision and mapping in manufacturing. In a recent project, Taggart has used the Turtlebot 2 robot to overcome motion blur, one of the biggest problems for feature-based simultaneous localization and mapping systems (SLAM). Motion blur, caused by movement or vibration of the camera, causes inaccuracies and location losses during SLAM.
The Turtlebot 2, sourced from Clearpath, was the base of a prototype indoor robotic vision system operating in warehouses, in which it was required to navigate through dynamic, consistently-lit, feature-dense, enclosed areas. Reducing motion blur will facilitate situation-agnostic improvement, enabling the robot to move faster and more accurately.