Precise pose graph localization with sparse point and lane features

Wu, Cong; Huang, Tianhe; Muffert, Maximilian; Schwarz, Tilo; Gräter, Johannes

doi:10.1109/iros.2017.8206264

Cited by 12 publications

(3 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A stereo camera and a laser scanner can be used to detect pole-like structures in urban environments [8]. Another option is using a range sensor for landmarks and a camera for road markings [25,26].…”

Section: Sensor Data Fusion Of Lidar Point Clouds and Camera Imagesmentioning

confidence: 99%

Investigation on Robustness of Vehicle Localization Using Cameras and LiDAR

Albrecht

Behre²,

Herrmann³

et al. 2022

Vehicles

View full text Add to dashboard Cite

Vehicle self-localization is one of the most important capabilities for automated driving. Current localization methods already provide accuracy in the centimeter range, so robustness becomes a key factor, especially in urban environments. There is no commonly used standard metric for the robustness of localization systems, but a set of different approaches. Here, we show a novel robustness score that combines different aspects of robustness and evaluate a graph-based localization method with the help of fault injections. In addition, we investigate the influence of semantic class information on robustness with a layered landmark model. By using the perturbation injections and our novel robustness score for test drives, system vulnerabilities or possible improvements are identified. Furthermore, we demonstrate that semantic class information allows early discarding of misclassified dynamic objects such as pedestrians, thus improving false-positive rates. This work provides a method for the robustness evaluation of landmark-based localization systems that are also capable of measuring the impact of semantic class information for vehicle self-localization.

show abstract

Section: Sensor Data Fusion Of Lidar Point Clouds and Camera Imagesmentioning

confidence: 99%

Investigation on Robustness of Vehicle Localization Using Cameras and LiDAR

Albrecht

Behre²,

Herrmann³

et al. 2022

Vehicles

View full text Add to dashboard Cite

show abstract

“…After getting the initial pose, pose tracking is performed to incrementally track the pose of a mobile manipulator. A graph-based pose tracking method using sparse point feature and lane marking is proposed by Wu et al [22]. Their approach performs with less than 0.50 m/0.41 deg localization error in an outdoor city environment.…”

Section: Pose Trackingmentioning

confidence: 99%

Hybrid Maps Enhanced Localization System for Mobile Manipulator in Harsh Manufacturing Workshop

Huang

Zhang

et al. 2020

IEEE Access

View full text Add to dashboard Cite

With excellent mobility and flexibility, mobile manipulators have great potential for loading and unloading tasks of numerical control machine tools (CNC) in manufacturing workshops. However, because of the rough and oily ground, dynamic obstacles and the convex plate of a CNC, harsh manufacturing workshop poses a huge challenge to the localization system of an autonomous mobile manipulator. To address the above problem, this paper presents a hybrid maps enhanced localization system which mainly consists of a global localization method and a pose tracking method. Hybrid maps including hybrid grid map, multiresolution likelihood fields (MLFs) and hybrid point map are constructed to efficaciously model the harsh environment and to improve localization performance. Our global localization method employs the convex hull sampling to spares dense Lidar data and the MLFs based branch and bound (BnB) search to speed up global search. To achieve real-time localization reliably and accurately, our pose tracking method seamlessly combines the BnB search and the adaptive Monte Carlo localization, and the Iterative Closest Point (ICP) based scan matching using the hybrid point map is adopted for higher accuracy. In addition, a distance filter improved by unscented transform is integrated into the pose tracking process to mitigate the influence of dynamic obstacles. The developed localization system is evaluated through different experiments including two weeks of loading and unloading tasks in a real manufacturing scenario, resulting in superior localization performance.

show abstract

“…To work on nonlinear systems, the extended Kalman filter 25 and the unscented Kalman filter 26 were developed. Optimization-based methods 27,28 have been demonstrated to have better performance than the Kalman filter and its variants on sophisticated nonlinear systems. Most localization methods suppose that all errors are Gaussian.…”

Section: Related Workmentioning

confidence: 99%

Lane-level localization system using surround-view cameras adaptive to different driving conditions

Qian

Fortelle

et al. 2020

International Journal of Advanced Robotic Systems

View full text Add to dashboard Cite

This article presents a lane-level localization system adaptive to different driving conditions, such as occlusions, complicated road structures, and lane-changing maneuvers. The system uses surround-view cameras, other low-cost sensors, and a lane-level road map which suits for mass deployment. A map-matching localizer is proposed to estimate the probabilistic lateral position. It consists of a sub-map extraction module, a perceptual model, and a matching model. A probabilistic lateral road feature is devised as a sub-map without limitations of road structures. The perceptual model is a deep learning network that processes raw images from surround-view cameras to extract a local probabilistic lateral road feature. Unlike conventional deep-learning-based methods, the perceptual model is trained by auto-generated labels from the lane-level map to reduce manual effort. The matching model computes the correlation between the sub-map and the local probabilistic lateral road feature to output the probabilistic lateral estimation. A particle-filter-based framework is developed to fuse the output of map-matching localizer with the measurements from wheel speed sensors and an inertial measurement unit. Experimental results demonstrate that the proposed system provides the localization results with submeter accuracy in different driving conditions.

show abstract

Precise pose graph localization with sparse point and lane features

Cited by 12 publications

References 16 publications

Investigation on Robustness of Vehicle Localization Using Cameras and LiDAR

Investigation on Robustness of Vehicle Localization Using Cameras and LiDAR

Hybrid Maps Enhanced Localization System for Mobile Manipulator in Harsh Manufacturing Workshop

Lane-level localization system using surround-view cameras adaptive to different driving conditions

Contact Info

Product

Resources

About