Map-Based Precision Vehicle Localization in Urban Environments

By Jesse Levinson, Michael Montemerlo, Sebastian Thrun. 2007

458 citations recorded by the source corpus.

Many urban navigation applications (e.g., autonomous navigation, driver assistance systems) can benefit greatly from localization with centimeter accuracy. Yet such accuracy cannot be achieved reliably with GPS-based inertial guidance systems, specifically in urban settings. We propose a technique for high-accuracy localization of moving vehicles that utilizes maps of urban environments. Our approach integrates GPS, IMU, wheel odometry, and LIDAR data acquired by an instrumented vehicle, to generate high-resolution environment maps. Offline relaxation techniques similar to recent SLAM methods [2, 10, 13, 14, 21, 30] are employed to bring the map into alignment at intersections and other regions of self-overlap. By reducing the final map to the flat road surface, imprints of other vehicles are removed. The result is a 2-D surface image of ground reflectivity in the infrared spectrum with 5cm pixel resolution. To localize a moving vehicle relative to these maps, we present a particle filter method for correlating LIDAR measurements with this map. As we show by experimentation, the resulting relative accuracies exceed that of conventional GPS-IMU-odometry-based methods by more than an order of magnitude. Specifically, we show that our algorithm is effective in urban environments, achieving reliable real-time localization with accuracy in the 10centimeter range. Experimental results are provided for localization in GPS-denied environments, during bad weather, and in dense traffic. The proposed approach has been used successfully for steering a car through narrow, dynamic urban roads.

Subfields: Aerospace Engineering, Artificial Intelligence, Electrical and Electronic Engineering

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Aerospace Engineering

In Artificial Intelligence

In Electrical and Electronic Engineering

Citations in the corpus

A Survey of Autonomous Driving: Common Practices and Emerging Technologies (2019, methodology): “GPS-IMU systems by themselves cannot be used for vehicle localization as they do not meet the performance criteria [109].”
Robust and Precise Vehicle Localization Based on Multi-Sensor Fusion in Diverse City Scenes (2017, methodology): “[2], [12] also relies on a pre-fused GNSS/IMU input.”
Computer Vision for Autonomous Vehicles: Problems, Datasets and State-of-the-Art (2017, background): “To alleviate this problem, Levinson et al. (2007) create a map only consisting of features that are very likely to be static.”
A Theoretical Foundation of Network Localization and Navigation (2018, background): “• Position: Node positions can be measured directly by, for example, a vision or Lidar sensor through a known local reference.”
Computing Systems for Autonomous Driving: State of the Art and Challenges (2020, background): “tion, distance estimation, edge detection, SLAM [17], [50], and high-definition (HD) map generation [18], [51]–[53].”
Efficient Surfel-Based SLAM using 3D Laser Range Data in Urban Environments (2018, unknown): “mapping systems mainly accomplish the estimation relying on feature-based solutions [34, 35], reduced map representations [14, 13], voxel grid-based methods [16], or point sub-sampling [30], which all effectively reduce the data used for alignment.”
Robust vehicle localization in urban environments using probabilistic maps (2010, methodology): “In fact, in order to enable autonomous driving, a localization system should be able to handle situations where the environment has changed since the map was created.”
A Survey of the State-of-the-Art Localization Techniques and Their Potentials for Autonomous Vehicle Applications (2018, methodology): “[21] proposed a solution integrating GPS, IMU, wheel odometry, and LiDAR to generate high-resolution”
Fast LIDAR localization using multiresolution Gaussian mixture maps (2015, methodology): “The pipeline can also be adapted to the probabilistic intensity maps of Levinson et al. [7].”
Simultaneous Localization and Mapping: A Survey of Current Trends in Autonomous Driving (2017, unknown): “Finally, an interesting approach, first proposed in [177] then extended in [178], focuses on building high-quality road surface maps using a 3D laser.”
Light-weight localization for vehicles using road markings (2013, methodology): “The most reliable solution to the vehicle localization problem thus far has been through the use of 3D lidar sensors, such as the Velodyne [10], which in addition to being expensive, are active sensors with their own set of challenges.”
Visual localization within LIDAR maps for automated urban driving (2014, unknown): “State-of-the-art methods [2], [3] use reflectivity measurements from 3D LIDAR scanners to create an orthographic map of ground-plane reflectivities.”
Towards fully autonomous driving: Systems and algorithms (2011, methodology): “Furthermore, by using offline SLAM to align multiple passes of the same environment[5], possibly separated in time by days or even months, it is possible to build an increasingly robust understanding of the world that can be then exploited for localization.”
The Architectural Implications of Autonomous Driving: Constraints and Acceleration (2018, background): “, precision at decimeter-level is needed to keep the vehicle staying in certain lanes [40]) and the spacial accessibility [5] to localize the vehicle (step 1b in Figure 1) for autonomous driving tasks.”
What could move? Finding cars, pedestrians and bicyclists in 3D laser data (2012, background): “This is witnessed by a cornucopia of publications in this area [1], [2], [3].”
Internet of vehicles in big data era (2018, methodology): “In [134], a precision vehicle localization method in urban environment is proposed based on high-resolution maps.”
Edge Computing for Autonomous Driving: Opportunities and Challenges (2019, methodology): “LiDAR is used for localization, relying heavily on a particle filter [6].”
L3-Net: Towards Learning Based LiDAR Localization for Autonomous Driving (2019, methodology): “Thrun [19, 20] propose a LiDAR intensity-based localization method.”