RGB-D and laser data fusion-based human detection and tracking for socially aware robot navigation framework

This paper proposes an effective socially aware navigation framework for mobile service robots in social environments. The proposed framework consists of three stages. In the first stage, RGB-D and laser data fusion-based human detection and tracking system is utilized to detect humans in the vicinity of the robot. In the second stage of framework, the extended personal space is modelled by using the human states including position and motion, and the relative motion between the human and the robot. In the third stage, the extended personal space is incorporated into the motion planning system and then a kinodynamic RRT motion planner is made use of to generate a legible trajectory of the mobile robot. The experimental results indicate that the proposed framework is able to ensure the safety of human, providing socially acceptable behaviors of the mobile service robot.