Efficient Integration of Octree Based Maps in Multi-Robot System

Abstract

The paper presents an approach for integration of three-dimensional world models (octree based maps) in multi-robot systems. The process of maps integration is based on ICP (Iterative Closest Point) algorithm that has been adapted and optimized to operate on octomaps. The presented method has been verified in numerous experiments both in simulation and with publicly available datasets captured from robots. The solution can be applied to various robotic applications such as underwater robots, aerial robots or robots equipped with manipulators - wherever 3D maps are needed and a task is performed by group of mobile robots.