Robust Self-Tuning Data Association for Geo-Referencing Using Lane Markings

  title={Robust Self-Tuning Data Association for Geo-Referencing Using Lane Markings},
  author={Miguel 'Angel Munoz-Ban'on and Jan-Hendrik Pauls and Haohao Hu and Christoph Stiller and Francisco A. Candelas and Fernando Torres},
  journal={IEEE Robotics and Automation Letters},
Localization in aerial imagery-based maps offers many advantages, such as global consistency, geo-referenced maps, and the availability of publicly accessible data. However, the landmarks that can be observed from both aerial imagery and on-board sensors is limited. This leads to ambiguities or aliasing during the data association. Building upon a highly informative representation (that allows efficient data association), this paper presents a complete pipeline for resolving these ambiguities… 

Figures and Tables from this paper



Accurate Global Trajectory Alignment using Poles and Road Markings

A novel geo-referencing approach to align trajectories to an aerial imagery using pole and road marking features using GNSS measurements from a low-cost GNSS receiver and yields trajectories which are accurately referenced to the used aerial imagery.

Localization in Aerial Imagery with Grid Maps using LocGAN

The LocGAN approach, a localization approach based on a geo-referenced aerial imagery and LiDAR grid maps, and a localization network (LocNet) is presented, which provides reliable global localization results.

DA-LMR: A Robust Lane Marking Representation for Data Association

DA-LMR (Delta-Angle Lane Marking Representation), a robust data representation that encodes how a curve changes in each point and includes this information in an additional dimension, thus providing a more detailed geometric structure description of the data.

Fusing Lidar Data and Aerial Imagery with Perspective Correction for Precise Localization in Urban Canyons

  • Jonghwi KimJinwhan Kim
  • Environmental Science
    2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)
  • 2019
This paper addresses a vehicle localization method that fuses aerial maps and lidar data in urban canyon environments where global positioning system (GPS) signals are inaccurate and a new method to correct perspective projection distortions is proposed.

Accurate Global Localization Using Visual Odometry and Digital Maps on Urban Environments

This paper describes a full solution to the estimation of the global position of a vehicle in a digital road map by means of visual information alone, based on a stereo platform used to estimate the motion trajectory of the ego vehicle and a map-matching algorithm, which will correct the cumulative errors of the vision-based motion information.

Global Localization on OpenStreetMap Using 4-bit Semantic Descriptors

This paper proposes an approach that builds upon publicly available map information from OpenStreetMap and turns them into a compact map representation that can be used for Monte Carlo localization, which requires to store only a tiny 4-bit descriptor per location and is still able to globally localize and track a vehicle.

Modeling Perceptual Aliasing in SLAM via Discrete–Continuous Graphical Models

Experimental results on standard benchmarking datasets show that the proposed technique compares favorably with state-of-the-art methods while not relying on an initial guess for optimization.

Global outer-urban navigation with OpenStreetMap

A probabilistic approach to autonomous robot navigation using data from OpenStreetMap that associates tracks from Open-StreeetMap with the trails detected by the robot based on its 3D-LiDAR data, using a Markov-Chain Monte-Carlo technique.

Monocular Localization in HD Maps by Combining Semantic Segmentation and Distance Transform

This work proposes using semantic segmentation on a monocular camera to localize directly in a HD map as used for automated driving, which combines lightweight, yet powerful HD maps with the simplicity of monocular vision and the flexibility of neural networks.

City-wide Street-to-Satellite Image Geolocalization of a Mobile Ground Agent

This work combines a neural network with a particle filter to achieve global position estimates for agents moving in GPS-denied environments, scaling efficiently to city-scale regions and reduces the final position estimation error by 64% compared to a state-of-the-art baseline from the literature.