Accurate surface normal estimation is critical for LiDAR-based tasks such as mapping, localization, and scene understanding. However, traditional 3D neighborhood selection methods often struggle with sparse point clouds due to uneven density and sensor limitations. In this paper, we propose a geometry-aware approach that leverages the structured 2D range view to select more consistent neighborhoods for normal estimation. By combining adaptive windowing in range space with distance-based filtering, our method provides robust and accurate normal vectors across both dense and sparse regions. We integrate this method into a lightweight SLAM framework and evaluate it using indoor LiDAR data from a Velodyne-16 sensor. The results demonstrate improved normal estimation and enhanced localization and mapping performance, highlighting the method’s effectiveness for indoor robotic applications.

Recently published datasets have been increasingly comprehensive with respect to their variety of simultaneously used sensors, traffic scenarios, environmental conditions, and provided annotations. However, these datasets typically only consider data collected by one independent vehicle. Hence, there is currently a lack of comprehensive, real-world, multi-vehicle datasets fostering research on cooperative applications such as object detection, urban navigation, or multi-agent SLAM. In this paper, we aim to fill this gap by introducing the novel LUCOOP dataset, which provides time-synchronized multi-modal data collected by three interacting measurement vehicles. The driving scenario corresponds to a follow-up setup of multiple rounds in an inner city triangular trajectory. Each vehicle was equipped with a broad sensor suite including at least one LiDAR sensor, one GNSS antenna, and up to three IMUs. Additionally, Ultra-Wide-Band (UWB) sensors were mounted on each vehicle, as well as statically placed along the trajectory enabling both V2V and V2X range measurements. Furthermore, a part of the trajectory was monitored by a total station resulting in a highly accurate reference trajectory. The LUCOOP dataset also includes a precise, dense 3D map point cloud, acquired simultaneously by a mobile mapping system, as well as an LOD2 city model of the measurement area. We provide sensor measurements in a multi-vehicle setup for a trajectory of more than 4 km and a time interval of more than 26 minutes, respectively. Overall, our dataset includes more than 54,000 LiDAR frames, approximately 700,000 IMU measurements, and more than 2.5 hours of 10 Hz GNSS raw measurements along with 1 Hz data from a reference station. Furthermore, we provide more than 6,000 total station measurements over a trajectory of more than 1 km and 1,874 V2V and 267 V2X UWB measurements. Additionally, we offer 3D bounding box annotations for evaluating object detection approaches, as well as highly accurate ground truth poses for each vehicle throughout the measurement campaign.

We present a novel interval-based visual-inertial LiDAR SLAM (i-VIL SLAM) method that solely assumes sensor errors to be bounded and propagates the error from the input sources to the estimated map and trajectory using interval analysis. The method allows us to restrict the solution set of the robot poses and the position of the landmarks to the set that is consistent with the measurements. If the error limits are not violated, it is guaranteed that the estimated set contains the true solution. The accumulation of the uncertainty is stabilized by anchoring poses derived from GNSS/INS data. Furthermore, for the first time we compare confidence ellipses determined by a classical SLAM graph optimization approach with the interval estimates of the robot poses provided by our method. In this work, we experimentally show that the marginal co-variances computed by the classical SLAM graph optimization are too overconfident and underestimate the uncertainty of the poses. While the 99.9 %-ellipsoids derived from the marginal covariances of the poses only enclose less than 64 % of the ground truth in the worst case, our method provides interval bounds for the pose parameters that enclose the ground truth for more than 96 % of all frames.

Improvements in sensor technologies as well as machine learning methods allow an efficient collection, processing and analysis of the dynamic environment, which can be used for detection and tracking of traffic participants. Current datasets in this domain mostly present a single view, making highly accurate pose estimation impossible due to occlusions. The integration of different, simultaneously acquired data allows to exploit and develop collaboration principles to increase the quality, reliability and integrity of the derived information. This work addresses this problem by providing a multi-view dataset, including 2D image information (videos) obtained by up to three cameras and 3D point clouds from up to five LiDAR sensors together with labels of the traffic participants in the scene. The measurements were conducted during different weather conditions on several days at a large junction in Hanover, Germany, resulting in a total duration of 145 minutes.

This paper presents a large-scale strip adjustment method for LiDAR mobile mapping data, yielding highly precise maps. It uses several concepts to achieve scalability. First, an efficient graph-based pre-segmentation is used, which directly operates on LiDAR scan strip data, rather than on point clouds. Second, observation equations are obtained from a dense matching, which is formulated in terms of an estimation of a latent map. As a result of this formulation, the number of observation equations is not quadratic, but rather linear in the number of scan strips. Third, the dynamic Bayes network, which results from all observation and condition equations, is partitioned into two sub-networks. Consequently, the estimation matrices for all position and orientation corrections are linear instead of quadratic in the number of unknowns and can be solved very efficiently using an alternating least squares approach.

Being regarded as essential for advanced driver assistance systems and autonomous driving, the research field of map generation from crowd-sourced vehicle information gained attention in the last decade. This paper introduces a novel approach for the derivation of street accurate road networks from such data. The method is applied to a real-world dataset of different accuracy level and is evaluated against a ground truth map. Furthermore, the results are compared to results of two state of the art algorithms.

This work aims at the inference of traffic networks from GPS trajectories. We perform geometry and topology reconstruction of the network in a multistep process. Our main contributions are the formulation of an explicit intersection model with a score function that accounts for consistency with the raw tracking data, as well as for a topology prior and the search for the best model by maximization of this score function using a Markov chain Monte Carlo sampler. We demonstrate the viability of our model-based approach with experiments on GPS data sets of varying size and data quality, followed by a comparison with results achieved by alternative, heuristic approaches.

Autonomous driving requires vehicle positioning with accuracies of a few decimeters. Typical low-cost GNSS sensors, as they are commonly used for navigation systems, are limited to an accuracy of several meters. Also, they are restricted in reliability because of outages and multipath effects. To improve accuracy and reliability, 3D features can be used, such as pole-like objects and planes, measured by a laser scanner. These features have to be matched to the reference data, given by a landmark map. If we use a nearest neighbor approach to match the data, we will likely get wrong matches, especially at positions with a low initial accuracy. To reduce the number of wrong matches, we use feature patterns. These patterns describe the spatial relationship of a specific number of features and are determined for every possible feature combination, separated in reference and online features. Given these patterns, the correspondences of the measured features can be determined by finding the corresponding patterns in the reference data. We acquired reference data by a high precision Mobile Mapping System. In an area of 2.8 km2we automatically extracted 1390 pole-like objects and 2006 building facades. A (second) vehicle equipped with an automotive laser scanner was used to generate features with lower accuracy and reliability. In every scan of the laser scanner we extracted landmarks (poles and planes) online. We then used our proposed feature matching to find correspondences. In this paper, we show the performance of the approach for different parameter settings and compare it to the nearest neighbor matching commonly used. Our experimental results show that, by using feature patterns, the rate of false matches can be reduced from about 80 % down to 20 %, compared to a nearest neighbor approach.

This paper presents a generative statistical approach to 3D building roof reconstruction from airborne laser scanning point clouds. In previous works bottom-up methods, e.g., points clustering, plane detection, and contour extraction, are widely used. Since the laser scanning data of urban scenes often contain extra structures and artefacts due to tree clutter, reflection from windows, water features, etc., bottom-up reconstructions may result in a number of incomplete or irregular roof parts.

We consider the problem of localizing a moving vehicle based on landmarks that were detected with a vehicle-mounted sensor. Landmarks are represented as points; correspondences of these points with the ones in a reference database are searched based on their geometric configurations. More specifically, we triangulate the landmark points and we match the obtained triangles with triangles in a reference database according to their geometric similarity. We maximize the number of triangle matches while considering the topological relations between different triangles, for example, if two triangles share an edge then the corresponding reference triangles must share an edge. Our method exploits that the observed points typically form a certain configuration: They appear at a limited distance from the vehicle's trajectory, thus the typical point pattern has a large extent in the driving direction and a relatively small lateral extent. This characteristic allows us to triangulate the observed point set such that we obtain a triangle strip (a sequence of triangles) in which each two consecutive triangles share one edge and each triangle connects three points that are relatively close to each other, that is, the triangle strip appropriately defines a neighborhood relationship for the landmarks. The adjacency graph of the triangles becomes a path; this allows for an efficient solution of our matching problem by dynamic programming. We present results of our method with data acquired with a mobile laser scanning system. The landmarks are objects of cylindric shape, for example, poles of traffic signs, which can be easily detected with the employed sensor. We tested the method with respect to its running time and its robustness when imposing different types of errors on the data. In particular, we tested the effect of non-rigid distortions of the observed point set, which are typically encountered during dead reckoning. Our matching approach copes well with such errors since it is based on local similarity measures of triangles, that is, we do not assume that a global non-rigid transformation between the observed point set and the reference point set exists.

Future driver assistance systems will require highly accurate positioning. One way to achieve this is by using on-board sensors to measure the relative location of landmarks for which the absolute coordinates are known.