GNSS-denied Ground Vehicle Localization for Off-road Environments with Bird’s-eye-view Synthesis

Abstract

Global localization is essential for the smooth navigation of autonomous vehicles. To obtain accurate vehicle states, on-board localization systems typically rely on Global Navigation Satellite System (GNSS) modules for consistent and reliable global positioning. However, in real-world scenarios, GNSS signals can be obstructed by natural or artificial barriers, leading to temporary system failures and degraded state estimation for autonomous vehicles.

On the other hand, off-road driving presents unique challenges for unmanned ground vehicles (UGVs) due to irregular terrain, leading to unstable surfaces for traversal that affect the accuracy of state estimation. Dense forests or canyons can block GNSS signals, hindering precise absolute positioning. Additionally, visual odometry performance may suffer due to the lack of distinct and reliable features necessary for accurate state estimation.

To address these challenges, we propose a novel learning-based method that synthesizes a local bird’s-eye-view (BEV) image of vehicle’s surrounding area by aggregating visual features from camera images. The proposed model combines a deformable attention-structured network with an image rendering head to generate a BEV image. The synthesized image is then matched with an aerial map for cross-view vehicle registration in GNSS-denied off-road environments.

Our method overcomes the limitations of visual inertial odometry (VIO) systems and the substantial storage requirements of image-retrieval-based localization strategies, which are susceptible to drift and scalability issues. Extensive real-world experimentation validates our method’s advancement over existing GNSS-denied visual localization methods, demonstrating notable enhancements in both localization accuracy and registration frequency. Furthermore, our method effectively reduces VIO drifts when integrated with an on-board VIO system via factor graph optimization.