Dense RGB-D-Inertial SLAM with Map Deformations

TL;DR

This paper introduces a real-time, dense RGB-D-inertial SLAM system that enhances robustness and accuracy in environment mapping by tightly integrating visual and inertial data, outperforming RGB-D-only methods especially during fast motions and low-texture scenarios.

Contribution

It presents the first tightly-coupled dense RGB-D-inertial SLAM system that jointly optimizes pose, velocity, biases, and gravity, with real-time GPU implementation.

Findings

More robust to fast motions

Handles low-texture environments better

Provides globally consistent dense reconstructions

Abstract

While dense visual SLAM methods are capable of estimating dense reconstructions of the environment, they suffer from a lack of robustness in their tracking step, especially when the optimisation is poorly initialised. Sparse visual SLAM systems have attained high levels of accuracy and robustness through the inclusion of inertial measurements in a tightly-coupled fusion. Inspired by this performance, we propose the first tightly-coupled dense RGB-D-inertial SLAM system. Our system has real-time capability while running on a GPU. It jointly optimises for the camera pose, velocity, IMU biases and gravity direction while building up a globally consistent, fully dense surfel-based 3D reconstruction of the environment. Through a series of experiments on both synthetic and real world datasets, we show that our dense visual-inertial SLAM system is more robust to fast motions and periods of low texture and low geometric variation than a related RGB-D-only SLAM system.