Reactive Collision Avoidance Using Real-Time Local Gaussian Mixture Model Maps
Authors:
- 1 October 2018
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 3545-3550
Abstract
In unknown, cluttered environments, robots require online real-time mapping and collision checking in order to navigate robustly. Discrete map representations are inefficient for collision checking as they are expensive in terms of memory and computation. This paper takes a probabilistic approach to local mapping by representing the environment as a Gaussian Mixture Model (GMM) and leverages its geometric properties to enable efficient collision checking given a time-parameterized trajectory. In contrast to current discretization-based methods, a GMM preserves geometric coverage of the environment without losing representation accuracy with varying map resolutions. We introduce a novel GMM local mapping algorithm that can be used with a single depth camera processed on a single CPU, and provide algorithms for collision avoidance given arbitrary trajectory representations. Finally, we provide experimentation results demonstrating safety, efficiency, and data coverage for real-time collision avoidance with a quadrotor navigating in a cluttered environment.Keywords
This publication has 11 references indexed in Scilit:
- On-Manifold GMM RegistrationIEEE Robotics and Automation Letters, 2018
- NanoMap: Fast, Uncertainty-Aware Proximity Queries with Lazy Search Over Local 3D Data2018
- A framework for efficient teleoperation via online adaptation2017
- Aggressive 3-D collision avoidance for high-speed navigation2017
- Computationally efficient information-theoretic exploration of pits and caves2016
- Online generation of collision-free trajectories for quadrotor flight in unknown cluttered environments2016
- MLMD: Maximum Likelihood Mixture Decoupling for Fast and Accurate Point Cloud Registration2015
- Stereo vision-based obstacle avoidance for micro air vehicles using disparity space2014
- Real-time navigation in 3D environments based on depth camera data2012
- Autonomous robot navigation using advanced motion primitives2009