Multi-task Gaussian Process Classification-based Collaborative Map Fusion Using Air-Ground Robotic System

Abstract

As one of the key building block of robotic system, current map fusion methods are usually constrained by using isomorphic modal of robotic systems or single type of sensors. They cannot works well in heterogeneous collaborative robotic systems that consists of aerial and ground robots. In this paper, we use heterogeneous robot systems consisting of unmanned ground vehicles (UGV) and unmanned aerial vehicle (UAV) to build occupancy grid maps that can be used for navigation. To fuse sensor data of different types, we propose a Collaborative Map Fusion algorithm based on Multi-task Gaussian Process Classification (MTGPC) using heterogeneous robotic systems. Our system is tested in real scenes and can achieve an accuracy of more than 70%. To our knowledge, this is the first work that can build the occupancy grid maps using sparse data points sampled from aerial images map and ground lidar map.