Artificial landmark for vision-based slam of water pipe rehabilitation robot

Abstract

Water supply has long history since urbanization has begun, and some of its pipes are too old to keep water quality. We have researched a robot system for water pipe rehabilitation. The system needs a SLAM (Simultaneous Localization and Mapping) algorithm as one of software components. However, conventional SLAM is not appropriate because water pipe is an environment requires special approaches. In this paper, we proposed an artificial landmark to solve the problem.