Improving Occupancy Grid Mapping via Dithering for a Mobile Robot Equipped with Solid-State LiDAR Sensors

Abstract

Occupancy grid maps are by far the most used spatial representation of the environment for robot navigation. This paper proposes a simple and effective way to improve the occupancy grid accuracy by superimposing a small oscillation to the robot motion when a predefined path is given. The method is especially suited for range sensors with long range capabilities but poor angular resolution. The innovative solid state LiDAR technology is an example of such sensor configuration and is used in this work for the experimental evaluation of the presented dithering technique. Experimental results quantitatively demonstrated that the proposed oscillating motion is effective especially in speeding up the detection of corridor like clearances in the environment.