Filtering and smoothing methods are used to produce an accurate estimate of the state of a time-varying system based on multiple observational inputs (data). Interest in these methods has exploded in recent years, with numerous applications emerging in fields such as navigation, aerospace engineering, telecommunications, and medicine. This compact, informal introduction for graduate students and advanced undergraduates presents the current state-of-the-art filtering and smoothing methods in a unified Bayesian framework. Readers learn what non-linear Kalman filters and particle filters are, how they are related, and their relative advantages and disadvantages. They also discover how state-of-the-art Bayesian parameter estimation methods can be combined with state-of-the-art filtering and smoothing algorithms. The book’s practical and algorithmic approach assumes only modest mathematical prerequisites. Examples include MATLAB computations, and the numerous end-of-chapter exercises include computational assignments. MATLAB/GNU Octave source code is available for download at www.cambridge.org/sarkka, promoting hands-on work with the methods.

贝叶斯滤波与平滑 (Bayesian filtering and smoothing)

This article provides a review of the production and uses of maps for autonomous driving and a synthesis of the opportunities and challenges. For many years, maps have helped human drivers make better decisions, and in the future, maps will continue to play a critical role in enabling safe and successful autonomous driving. There are, however, many technical, societal, economic, and political challenges to mapping that remain unresolved. While fully autonomous driving may be some distance in the future, intermediate steps to realize the technology can be taken. These include developing an efficient and reliable storage and dissemination infrastructure, defining minimum data quality requirements, and establishing an international mapping standard. The article closes with 11 open research challenges for mapping for autonomous driving.

Mapping for Autonomous Driving: Opportunities and Challenges

Point-cloud segmentation of 3D LiDAR scans is an important preprocessing task for autonomous vehicles in on-road and especially in off-road scenarios. Clustering point measurements with the same properties into multiple homogeneous regions is a challenging task due to an uneven sampling density and lack of explicit structural information. This paper presents a novel technique to achieve a robust and fast point-cloud segmentation using the characteristic intrinsic sensor pattern. This pattern is characterized by the mounting position of each laser diode. A structured mesh graph is created by taking the beam calibration and the chronology of incoming data packets into account. The proposed graph-based, multi-pass point segmentation algorithm compares this pattern with a flat-world model to detect discontinuities and to set label attributes such as obstacle or free space for each vertex. Furthermore, we directly detect missing measurements and therefore generate artificial vertices considering the laser beam intrinsics. Finally, a region-growing algorithm is applied in order to obtain cohesive objects. Experimental results show that we achieve a reliable overall performance and a good trade-off between segmentation quality and runtime of 15ms in rough terrain as well as suburban areas.

Fast Multi-Pass 3D Point Segmentation Based on a Structured Mesh Graph for Ground Vehicles

An important requirement for safe autonomous driving is the perception of dynamic and static objects. In urban scenarios, there exist hundreds of potential obstacles. Therefore, it is crucial to have a fast and accurate fitting method which is a key step for many tracking algorithms. In this paper, we demonstrate an orientation corrected bounding box fit based on the convex hull and a line creation heuristic. Our method is capable of fitting hundreds of objects in less than 10 ms and involves only few tuning parameters. Furthermore, orientation estimated through the dynamics of the object can be used to improve the fitting result. Real-world experiments have proven the robustness and effectiveness of our method.

An Orientation Corrected Bounding Box Fit Based on the Convex Hull under Real Time Constraints

3D object detection and classification are crucial tasks for perception in Autonomous Driving (AD). To promptly and correctly react to environment changes and avoid hazards, it is of paramount importance to perform those operations with high accuracy and in real-time. One of the most widely adopted strategies to improve the detection precision is to fuse information from different sensors, like e.g. cameras and LiDAR. However, sensor fusion is a computationally intensive task, that may be difficult to execute in real-time on an embedded platforms. In this paper, we present a new approach for LiDAR and camera fusion, that can be suitable to execute within the tight timing requirements of an autonomous driving system. The proposed method is based on a new clustering algorithm developed for the LiDAR point cloud, a new technique for the alignment of the sensors, and an optimization of the Yolo-v3 neural network. The efficiency of the proposed method is validated comparing it against state-of-the-art solutions on commercial embedded platforms.

Real-Time clustering and LiDAR-camera fusion on embedded platforms for self-driving cars

Robust object detection is one of the key tasks for autonomous vehicles. Clustering is the fundamental step for extracting objects from 3D point clouds. We propose a fast and efficient algorithm to cluster 3D point clouds provided by modern LiDAR sensors. The clustering is based on graph theory and local contextual information. Our method encodes weights of graph edges by adopting perceptual laws based on the intrinsic sensor beam pattern. This significantly increases the robustness of the segmentation process. It allows a point-wise clustering even at challenging distances and viewing angles as well as occlusions. For the sake of speed, the clustering pipeline is separated into vertical and horizontal clustering. Therefore, we split the graph into multiple vertical and horizontal line graphs which are processed in parallel. Finally, the partitioned results are merged into coherent objects using a breadth-first search algorithm. Experiments in different suburban datasets have demonstrated that our proposed method outperforms other state of the art methods, especially in complex scenes. A quantitative comparison between our method and other representative clustering methods proves the efficiency and the effectiveness of our work.

Fast Dual Decomposition based Mesh-Graph Clustering for Point Clouds

Reliable collision avoidance is one of the main requirements for autonomous driving. Hence, it is important to correctly estimate the states of an unknown number of static and dynamic objects in real-time. Here, data association is a major challenge for every multi-target tracker. We propose a novel multi-target tracker called Greedy Dirichlet Process Filter (GDPF) based on the non-parametric Bayesian model called Dirichlet Processes and the fast posterior computation algorithm Sequential Updating and Greedy Search (SUGS). By adding a temporal dependence we get a real-time capable tracking framework without the need of a previous clustering or data association step. Real-world tests show that GDPF outperforms other multi-target tracker in terms of accuracy and stability.

The Greedy Dirichlet Process Filter -An Online Clustering Multi-Target Tracker

In this paper, we present a simultaneous localization and mapping framework that combines filter-based road course tracking and GraphSLAM for localization and mapping in unstructured and rural areas. Road perception plays a crucial role, especially in areas without road markings, precise point positioning or HD-maps. In order to improve vehicle localization, we detect and track the road course. The road is modeled with a novel B-Spline measurement model and tracked with an Unscented Kalman Filter. LiDAR measurements from a Velodyne sensor are preprocessed, clustered and used as input. In addition, the road network from OpenStreetMap (OSM) is utilized to improve robustness in road detection and tracking. To correct a global vehicle position offset, we estimate the transformation between the tracked road course and the road network from OSM with an iterative closest point algorithm. For road course mapping as well as offset correction and smoothing of the vehicle trajectory, all information is accumulated and fused using GraphSLAM. Evaluation of our method is performed by comparing real data, collected with an experimental vehicle in an unstructured and urban area. Localization results are compared to a high-precision RTK INS/GNSS system. The results show that our method is able to capture the road robustly and to improve the global vehicle position under challenging environmental conditions.

Unstructured Road SLAM using Map Predictive Road Tracking

This paper proposes fast and novel methods to jointly estimate the target’s unknown 3D shape and dynamics. Measurements are noisy and sparsely distributed 3D points from a light detection and ranging (LiDAR) sensor. The methods utilize non-uniform rational B-splines (NURBS) surfaces to approximate the target’s shape. One method estimates Cartesian scaling parameters of a NURBS surface, whereas the second method estimates the corresponding NURBS weights, too. Major advantages are the capability of estimating a fully 3D shape as well as the fast processing time. Real-world evaluations with a static and dynamic vehicle show promising results compared to state-of-the-art 3D extended target tracking algorithms.

Patrick Burger

Papers

Fast Multi-Pass 3D Point Segmentation Based on a Structured Mesh Graph for Ground Vehicles

Fast Dual Decomposition based Mesh-Graph Clustering for Point Clouds

The Greedy Dirichlet Process Filter -An Online Clustering Multi-Target Tracker

Unstructured Road SLAM using Map Predictive Road Tracking

Fast 3D Extended Target Tracking using NURBS Surfaces