AtomMap: A probabilistic amorphous 3D map representation for robotics and surface reconstruction

  title={AtomMap: A probabilistic amorphous 3D map representation for robotics and surface reconstruction},
  author={David Fridovich-Keil and Erik Nelson and Avideh Zakhor},
  journal={2017 IEEE International Conference on Robotics and Automation (ICRA)},
We present a new 3D probabilistic occupancy map representation for robotics applications by relaxing the commonly-assumed constraint that space must be perfectly tessellated. We replace the regular structure of 3D grids with an unstructured collection of non-overlapping, equally-sized spheres, which we call “atoms”. Abandoning the grid structure allows a more accurate representation of space directly tangent to surfaces, which facilitates a number of applications such as high fidelity surface… 

Figures and Tables from this paper

Autonomous Navigation in Unknown Environments with Sparse Bayesian Kernel-based Occupancy Mapping
This paper provides an online training algorithm, updating the sparse Bayesian map incrementally from streaming range data, and an efficient collision-checking method for general curves, representing potential robot trajectories, and the effectiveness of the mapping and collision checking algorithms is evaluated in tasks requiring autonomous robot navigation in unknown environments.
Autonomous Navigation in Unknown Environments using Sparse Kernel-based Occupancy Mapping
A new map representation is proposed, in which occupied and free space are separated by the decision boundary of a kernel perceptron classifier, and an online training algorithm is developed that maintains a very sparse set of support vectors to represent obstacle boundaries in configuration space.
DCT Maps: Compact Differentiable Lidar Maps Based on the Cosine Transform
DCT maps represent lidar measurements significantly more accurate than grid maps, Gaussian process occupancy maps, and Hilbert maps, all with the same memory requirements, as demonstrated in the real-world experiments.
Towards Cooperative SLAM for Low-Cost Biomimetic Robots
A collection of estimation and control techniques allowing teams of millirobots to localize within and map complex, unstructured environments, including low-cost millirobot team design, motion modeling, cooperative state estimation, and mapping.
Real-Time Compact Environment Representation for UAV Navigation
A novel vertical strip extraction algorithm is proposed to analyze the probability density function characteristics of the normalized disparity value and segment the obstacles through an adaptive size sliding window and represent the obstacle surfaces as polygonal prism profiles while minimizing the redundant obstacle information.
Probabilistic Online Learning of Appearance and Structure for Robotics
This dissertation presents a probabilistic online self-learning framework to alleviate the dependency in robotic visual learning by leveraging structural priors and an online framework for continuous spatial mapping using Gaussian Processes.
Integration of Constructive Solid Geometry and Boundary Representation (CSG-BRep) for 3D Modeling of Underground Cable Wells from Point Clouds
The characteristics of 3D terrestrial lidar point cloud data are utilized to build a CSG-BRep 3D model of underground cable wells, whose spatial topological relationship is fully considered and meets the current production demands.
Try to Start It! The Challenge of Reusing Code in Robotics Research
  • E. Cervera
  • Computer Science
    IEEE Robotics and Automation Letters
  • 2019
Docker, a Linux container technology, is proposed to turn the source code repositories into executable images, that can be run and tested locally, in an isolated environment, without the need of a costly integration with the host system.
Automated Damage Assessment of Critical Infrastructure Using Online Mapping Technique with Small Unmanned Aircraft Systems
This paper proposes a data collection, detection and assessment system, using a sUAS equipped with a Lidar and a camera, that provides a solution in fast post-disaster response and assists human responders in damage investigation.
Towards Automated Post-Disaster Damage Assessment of Critical Infrastructure with Small Unmanned Aircraft Systems
This method provides new capabilities for automated damage analysis with sUAS as a key enabling technology for documenting damage in critical infrastructure to facilitate the post-disaster inspection and recovery of critical infrastructures.


OctoMap: an efficient probabilistic 3D mapping framework based on octrees
An open-source framework to generate volumetric 3D environment models based on octrees and uses probabilistic occupancy estimation that represents not only occupied space, but also free and unknown areas and an octree map compression method that keeps the 3D models compact.
Signed Distance Fields: A Natural Representation for Both Mapping and Planning
This paper proposes to combine a commonly-used representation in computer graphics and surface reconstruction, projective Truncated Signed Distance Field (TSDF), with a representation frequently used for collision checking and collision costs in planning, Euclidean Signed distance Field (ESDF), and validate this combined approach in simulation.
Planning 3-D Path Networks in Unstructured Environments
Using a priori aerial data scans of forested environments, a network of free space bubbles forming safe paths within environments cluttered with tree trunks, branches and dense foliage is computed.
Multi-volume occupancy grids: An efficient probabilistic 3D mapping model for micro aerial vehicles
This paper suggests a new approach to 3D mapping using a multi-volume occupancy grid, or MVOG, which outperform existing probabilistic3D mapping methods in terms of memory usage, due to the fact that observations are grouped together into continuous vertical volumes to save space.
Multi-Level Surface Maps for Outdoor Terrain Mapping and Loop Closing
This paper proposes a new representation denoted as multi-level surface maps (MLS maps) which allows to store multiple surfaces in each cell of the grid and is well-suited for representing large-scale outdoor environments.
Non-cubic occupied voxel lists for robot maps
  • J. RydeM. Brünig
  • Mathematics
    2009 IEEE/RSJ International Conference on Intelligent Robots and Systems
  • 2009
An alternative to the conventional quantization for occupied voxel lists in both 2D and 3D is presented and the D3 or face centred cubic lattice is highlighted for its lower quantization error, lower rotation variability and higher order rotations.
Terrain mapping for a roving planetary explorer
The authors are prototyping a legged vehicle, the Ambler, for an exploratory mission on another planet, conceivably Mars, where it is to traverse uncharted areas and collect material samples and present an algorithm for constructing an elevation map from a single range image.
Lattice occupied voxel lists for representation of spatial occupancy
  • J. RydeM. Brünig
  • Mathematics
    2010 IEEE/RSJ International Conference on Intelligent Robots and Systems
  • 2010
An algorithm for generic quantization to any low dimensional lattice is presented allowing this analysis to be easily extended to other 3D lattices of interest.
Information-theoretic mapping using Cauchy-Schwarz Quadratic Mutual Information
We develop a computationally efficient control policy for active perception that incorporates explicit models of sensing and mobility to build 3D maps with ground and aerial robots. Like previous
Robot spatial perception by stereoscopic vision and 3D evidence grids
A new program that derives a dense three-dimensional evidence grid representation of a robot’s surroundings from wide-angle stereoscopic images seems abundantly adequate for very reliable navigation of freely roaming mobile robots, and plausibly adequate for shape identification of objects bigger than 10 centimeters.