TY - CONF
T1 - Visual Support for Navigation in the Polly System
AU - Horswill, Ian
N1 - Funding Information:
Navigation is one of the most basic problems in robotics. Since the ability to safely move about the world is a prerequisite of most other activities, navigation has received a great deal of attention in the AI, robotics, and computer vision communities. One of the limiting factors in the design of current navigation systems, as with many other robotic systems, has been the availability of reliable sensor data. Most systems have relied on the use of sonar data [7][9], or on vision [10][3][5][6][13][1]. In all cases, the unreliability of the available sensor data was a major concern in the research. Somer esearchers have even avoided the use of sensor data entirely in favor of precompiled maps [11]. In this paper, I will discuss a very simple vision-based corridor following system which is in day-to-day use on Polly, a low cost mobile robot. The system is notable in that it is very fast (14 frames per second in the current 1Support for this research was provided in part by the University Research Initiative under Office of Naval Research contract N00014-86-K-0685a,n d in part by the Advanced Research Projects Agencyu nder Office of Naval Research contract N00014-85-K-0124T. hanks to Steve White for reading an early draft of this paper.
Publisher Copyright:
Copyright © 1992, AAAI (www.aaai.org). All rights reserved.
PY - 1992
Y1 - 1992
N2 - In this paper I will discuss a system which uses vision to guide a mobile robot through corridors and freespace channels. Thesystem runs in an unmodified office environmentin the presence of both static and dynamic obstacles (e.g. people). The system is also amongthe simplest, most effective, and best tested systems for vision-based navigation to date. The performance of the system is dependent on an analysis of both the task and the environment in which it is performed. I will describe both of these and discuss howthey simplify the computational problemsfacing the robot, as well as the performanceof the resulting system. Finally, I will briefly discuss workin progress regarding place recognition 1 and point-to-point navigation.
AB - In this paper I will discuss a system which uses vision to guide a mobile robot through corridors and freespace channels. Thesystem runs in an unmodified office environmentin the presence of both static and dynamic obstacles (e.g. people). The system is also amongthe simplest, most effective, and best tested systems for vision-based navigation to date. The performance of the system is dependent on an analysis of both the task and the environment in which it is performed. I will describe both of these and discuss howthey simplify the computational problemsfacing the robot, as well as the performanceof the resulting system. Finally, I will briefly discuss workin progress regarding place recognition 1 and point-to-point navigation.
UR - http://www.scopus.com/inward/record.url?scp=85168701892&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85168701892&partnerID=8YFLogxK
M3 - Paper
AN - SCOPUS:85168701892
SP - 62
EP - 67
T2 - 1992 AAAI Fall Symposium on Applications of AI to Real-World Autonomous Mobile Robots
Y2 - 23 October 1992 through 25 October 1992
ER -