GOMSF: Graph-Optimization Based Multi-Sensor Fusion for robust UAV Pose estimation

Abstract

Achieving accurate, high-rate pose estimates from proprioceptive and/or exteroceptive measurements is the first step in the development of navigation algorithms for agile mobile robots such as Unmanned Aerial Vehicles (UAVs). In this paper, we propose a decoupled Graph-Optimization based Multi-Sensor Fusion approach (GOMSF) that combines generic 6 Degree-of-Freedom (DoF) visual-inertial odometry poses and 3 DoF globally referenced positions to infer the global 6 DoF pose of the robot in real-time. Our approach casts the fusion as a real-time alignment problem between the local base frame of the visual-inertial odometry and the global base frame. The alignment transformation that relates these coordinate systems is continuously updated by optimizing a sliding window pose graph containing the most recent robot's states. We evaluate the presented pose estimation method on both simulated data and large outdoor experiments using a small UAV that is capable to run our system onboard. Results are compared against different state-of-the-art sensor fusion frameworks, revealing that the proposed approach is substantially more accurate than other decoupled fusion strategies. We also demonstrate comparable results in relation with a finely tuned Extended Kalman Filter that fuses visual, inertial and GPS measurements in a coupled way and show that our approach is generic enough to deal with different input sources in a straightforward manner.Videohttps://youtu.be/GIZNSZ2soL8

SIMSF: A Scale Insensitive Multi-Sensor Fusion Framework for Unmanned Aerial Vehicles Based on Graph Optimization

“…Merfels [18] proposed a sliding window method to fuse VIO and GPS measurements by optimizing the states in the window, and also proposed a corresponding marginalization method. Similar work can also be found in GOMSF [20], but the difference between GOMSF and [18] is that GOMSF estimated the transformation between the local frame of the VIO and the global frame to achieve a higher frequency estimation output rather than estimating the current state directly. In [19], [21], a prior map was needed for localization with a weak GPS environment.…”

Mitigation of Odometry Drift with a Single Ranging Link in GNSS-limited Environments

“…To estimate the vehicle's trajectory with respect to a global frame with reduced estimation error, absolute position measurements obtained with GNSS-based positioning can be fused with the on-board measurements (obtained with vision or visual-inertial odometry). In [1], [2], [3], a framework to fuse visual-inertial odometry and GPS positions is proposed, re-formulating the data fusion problem as a frame alignment problem by decoupling locally-referred odometry measurements and globally-referred poses. Although this approach can effectively mitigate the drift of the visual-inertial system, it requires a constantly available and reliable GNSS-based positioning, which is challenging to fulfill in many environments.…”

Cooperative swarm localization and mapping with inter-agent ranging

“…In these figure, we cannot see a large improvement of absolute positioning with the proposed method, which is expected since our approach exploits interagent range measurements that only provide relative information to the system. To mitigate the drifts in global position estimates, we need to use absolute measurements, such as absolute position measurements obtained with GNSS-based systems [19] and ranging links to a base station (anchor point) as proposed in [20] and [21].…”