MRFMap: Online Probabilistic 3D Mapping using Forward Ray Sensor Models

TL;DR

This paper introduces MRFMap, an online probabilistic 3D mapping method that explicitly models sensor ray formation and occlusions using Markov Random Fields, leading to more accurate occupancy maps from noisy sensor data.

Contribution

The paper presents a novel framework that explicitly models sensor ray formation and occlusions with MRFs, improving 3D map accuracy over traditional methods.

Findings

Higher fidelity occupancy maps demonstrated on simulated data.

Effective handling of noisy sensor data through learned noise characteristics.

New metric for evaluating probabilistic volumetric maps.

Abstract

Traditional dense volumetric representations for robotic mapping make simplifying assumptions about sensor noise characteristics due to computational constraints. We present a framework that, unlike conventional occupancy grid maps, explicitly models the sensor ray formation for a depth sensor via a Markov Random Field and performs loopy belief propagation to infer the marginal probability of occupancy at each voxel in a map. By explicitly reasoning about occlusions our approach models the correlations between adjacent voxels in the map. Further, by incorporating learnt sensor noise characteristics we perform accurate inference even with noisy sensor data without ad-hoc definitions of sensor uncertainty. We propose a new metric for evaluating probabilistic volumetric maps and demonstrate the higher fidelity of our approach on simulated as well as real-world datasets.