A Sampling-based Next-Best-View Path Planner for Environment Exploration

Abstract

A new path planner to explore the surrounding environment algorithm is presented for the aerial robot. This algorithm combines Next-Best-View (NBV) sampling and frontier exploration together to find unknown large areas. We first compute the nearby area, namely the frontier area, which can guide the exploration direction for the robot roughly, then a node tree representing the NBV of the robot is generated for the robot to follow. Next, a continuous trajectory-generated method is proposed where a B-spline curve representing the robot trajectory can follow these selected viewpoints in the history tree. We evaluate the effect of our algorithm in a simulated environment.