Abstract-Because of the difficulty of interpreting laser data in a meaningful way, safe navigation in vegetated terrain is still a daunting challenge. In this paper, we focus on the segmentation of ladar data using local 3-D point statistics into three classes: clutter to capture grass and tree canopy, linear to capture thin objects like wires or tree branches, and finally surface to capture solid objects like ground terrain surface, rocks or tree trunks. We present the details of the method proposed, the modifications we made to implement it on-board an autonomous ground vehicle. Finally, we present results from field tests using this rover and results produced from different stationary laser sensors.

No abstract

Abstract

Most research on 3-D object classification and recognition focuses on recognition of objects in 3-

As part of our long term research interests in examining the technological requirements for a shopping mall robot, we performed a short pilot study during the Christmas holidays to identify the social interaction dynamics for Neel,a wheeled mobile robot which interacts using its on-site and online presence. During the pilot study we found that a range of the robot's interaction capabilities were mostly unused due to the relatively short interactions users had with the robot to fulfill their informational requirements like the movie show-times or apparel deals. Since, its challenging to educate and inform a diverse mass of users about the robot's functionality, we decided to divide the research roadmap in stages where in the first stage the users would learn the value of the robot's capabilities by the repeated short-time interactions and over the long term more users would register for the robot's services. Hence, identifying the temporal and episodic characteristics of the interactions are perceived important to match the expectations and privacy concerns of the users. We also identified that while non-interactively delivering shopping related information through a web application is relatively easier, doing it actively through the robot can be probed as advertisements very easily by human participants and negate the user experience we try to deliver. We report some key technological advances we made through our field trial and set forth the goals.