SLAM Mapping with a Human Following Robot Using Depth Camera
Department of Computer Science and Engineering, The Chinese University of Hong Kong
Abstract
This research explored an alternative way of conducting SLAM mapping in indoor environments. Instead of using a remote controller to drive the robot car, the user leads the robot through an unknown area while the system performs mapping.
The project combined SLAM, navigation, and vision-based object following. Some parts were built from the ground up, while others were implemented from existing libraries and algorithms. The robot car was based on the Robot Operating System (ROS), an open-source framework for coordinating low-level hardware control with higher-level algorithmic behavior.
The functionality of SLAM and navigation was tuned around the core principles of the underlying algorithms. Vision-based object tracking was also explored through multiple approaches, with their effectiveness analyzed and discussed. Once those individual pieces were integrated, the robot was able to follow a person while generating a map of an unknown indoor environment.