Atlas Fusion -- Modern Framework for Autonomous Agent Sensor Data Fusion

TL;DR

Atlas Fusion is an open-source, scalable framework that integrates diverse sensor data for autonomous vehicles to create a comprehensive 3D environment model, supporting decision-making and planning.

Contribution

It introduces a universal, ROS-compatible sensor fusion platform that combines multiple sensor types into a flexible, extendable environment modeling system for autonomous robots.

Findings

Supports fusion of RGB, thermal cameras, LiDAR, IMU, GNSS

Provides a complete pipeline from data loading to visualization

Open-source implementation available for community use

Abstract

In this paper, we present our software sensor fusion framework for self-driving cars and other autonomous robots. We have designed our framework as a universal and scalable platform for building up a robust 3D model of the agent's surrounding environment by fusing a wide range of various sensors into the data model that we can use as a basement for the decision making and planning algorithms. Our software currently covers the data fusion of the RGB and thermal cameras, 3D LiDARs, 3D IMU, and a GNSS positioning. The framework covers a complete pipeline from data loading, filtering, preprocessing, environment model construction, visualization, and data storage. The architecture allows the community to modify the existing setup or to extend our solution with new ideas. The entire software is fully compatible with ROS (Robotic Operation System), which allows the framework to cooperate with other ROS-based software. The source codes are fully available as an open-source under the MIT license. See https://github.com/Robotics-BUT/Atlas-Fusion.