Lane-level Map Research Articles

A lane-level localization method via the lateral displacement estimation model on expressway

The gradual proliferation of high-definition (HD) maps has played a pivotal role in the advancement of intelligent vehicles. However, a considerable number of vehicles still face limitations in harnessing HD maps for high-precision navigation due to the absence of lane-level localization information. To address this challenge, this paper presents a lightweight lane-level localization method based on ubiquitous low-cost on-board vehicle cameras and Global Navigation Satellite System (GNSS) receivers. Initially, the line anchor-based feature refinement network is combined with the DeepSort-based lane tracking algorithm to detect and track the lanes in the image sequences. Subsequently, a lateral displacement estimation (LDE) model is proposed to establish the relationship between the vehicle's lateral displacement value in the ego lane and the slope of the lanes on both sides. Moreover, a lateral motion estimation (LME) method is employed to track the motion of the vehicle across multiple lanes. Finally, a lane-level map matching method is proposed based on the GNSS data and lateral displacement information. In the experimental section, three distinct types of dashcams have been installed in a vehicle, and high-precision positioning equipment is utilized as the ground truth for algorithm validation. The experimental results indicate that the proposed method can be effectively deployed in vehicles, enabling lane-level vehicle positioning on expressway. The code has been open-sourced and can be accessed at https://codeocean.com/capsule/6640813/.

Lane-Level Map Generation and Management Framework Using Connected Car Data

This study proposes a lane-level map generation and management framework using connected sensor data to reduce the manpower and time required for producing and updating high-definition (HD) maps. Unlike previous studies that relied on the onboard processing capabilities of vehicles to collect map-constructing elements, this study offloads computing for map generation to the cloud, assigning vehicles solely the role of transmitting sensor data. For efficient data collection, we divide the space into a grid format to define it as a partial map and establish the state of each map and its transition conditions. Lastly, tailored to the characteristics of the road elements composing the map, we propose an automated map generation technique and method for selectively collecting data. The map generation method was tested using data collected from actual vehicles. By transmitting images with an average size of 350 KB, implementation was feasible even with the current 5G upload bandwidth. By utilizing 12,545 elements, we were able to achieve a position accuracy and regression RMSE of less than 0.25 m, obtaining 651 map elements to construct the map. We anticipate that this study will help reduce the manpower and time needed for deploying and updating HD maps.

Targeting Lane-Level Map Matching for Smart Vehicles: Construction of High-Definition Road Maps Based on GIS

The development of smart vehicles has increased the demand for high-definition road maps. However, traditional road maps for vehicle navigation systems are not sufficient to meet the requirements of intelligent vehicle systems (e.g., autonomous driving). The present work comes up with a method of generating high-definition map models based on the geographic information system (GIS). A systematic map construction framework including the road layer, intersection connection layer, and lane layer is proposed based on the GIS database. Specifically, the constrained Delaunay triangular network method is applied to extract road layer network models, which are then used as linear reference networks to construct lane-level road maps. To further examine the feasibility of the proposed framework, a field experiment is then conducted to build a high-definition road map. Furthermore, a lane-level map matching test is conducted in the constructed road map using the trajectory data collected from a probe vehicle. The results show that the proposed method provides an efficient way of extracting lane-level information from urban road networks and can be applied for lane-level map matching with good performance.

Lane-Level Road Map Construction considering Vehicle Lane-Changing Behavior

In recent years, the construction of lane-level road maps has received extensive attention from industry and academia. It has been widely studied because this kind of map provides the foundation for much research, such as high-precision navigation, driving behavior analysis, and traffic analysis. Trajectory-based crowd-mapping is an emerging approach to lane-level map construction. However, the major problem is that existing methods neglect modeling the trajectory distribution in the longitudinal direction of the road, which significantly impacts precision. Thus, this article proposes a two-stage method based on vehicle lane-changing behavior to model the road’s lateral and longitudinal trajectory distributions simultaneously. In the first stage, lane-changing behaviors are extracted from vehicle trajectories. In the second stage, the lane extraction model is established using the weighted constrained Gaussian mixture model and hidden Markov model to estimate lane parameters (e.g., lane counts and lane centerline) on each road cross section. Then accurate and continuous lane centerlines can be constructed accordingly. The proposed method is verified using vehicle trajectory data collected from the crowdsourced platform named Mapillary. The results show that the proposed method can construct lane-level road information satisfactorily.

TrajGAT: A map-embedded graph attention network for real-time vehicle trajectory imputation of roadside perception

With the increasing deployment of roadside sensors, vehicle trajectories can be collected for driving behavior analysis and vehicle-highway automation systems. However, due to dynamic occlusions, vehicles are often lost from the view of roadside sensors, strongly affecting the data availability. To address this issue, we propose a novel deep learning framework to impute missing Trajectory data called map-embedded Graph ATtention network (TrajGAT). The framework splits the problem into two subtasks, a trajectory forecasting task based on historical data and an imputation task based on the forecasting results and real-time incomplete observational data. Temporal features are extracted and fused following an encoder-decoder architecture. To model dynamic spatial patterns, we introduce a sparse heterogeneous graph construct technique via vectorized lane-level map and a rule-based graph attention network, which can effectively capture remote dependencies and highlight key adjacency relationships. Numerical experiments based on the Argoverse imputation dataset and Lyft dataset are conducted to compare our TrajGAT and other state-of-the-art models. The results indicate that our model performs best based on various evaluation indicators and has strong robustness with different missing trajectory rates. The learned dynamic interaction can further help achieve a better understanding of the spatiotemporal dependency of vehicles in complex traffic scenarios.

FusionLane: Multi-Sensor Fusion for Lane Marking Semantic Segmentation Using Deep Neural Networks

Effective semantic segmentation of lane marking is crucial for construction of high-precision lane level maps. In recent years, a number of different methods for semantic segmentation of images have been proposed. These methods concentrate mainly on analysis of camera images, due to limitations with the sensor itself, and thus far, the accurate three-dimensional spatial position of the lane marking could not be obtained, which hinders lane level map construction.This article proposes a lane marking semantic segmentation method based on LIDAR and camera image fusion using a deep neural network. In the approach, the object of the semantic segmentation is a bird’s-eye view converted from a LIDAR points cloud instead of an image captured by a camera. First, the DeepLabV3+ network image segmentation method is used to segment the image captured by the camera, and the segmentation result is then merged with the point clouds collected by the LIDAR as the input of the proposed network. A long short-term memory (LSTM) structure is added to the neural network to assist the network in semantic segmentation of lane markings by enabling use of time series information. Experiments on datasets containing more than 14,000 images, which were manually labeled and expanded, showed that the proposed method provides accurate semantic segmentation of the bird’s-eye view LIDAR points cloud. Consequently, automation of high-precision map construction can be significantly improved. Our code is available at <uri xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">https://github.com/rolandying/FusionLane</uri> .

A Lane-Level Road Marking Map Using a Monocular Camera

The essential requirement for precise localization of a self-driving car is a lane-level map which includes road markings (RMs). Obviously, we can build the lane-level map by running a mobile mapping system (MMS) which is equipped with a high-end 3D LiDAR and a number of high-cost sensors. This approach, however, is highly expensive and ineffective since a single high-end MMS must visit every place for mapping. In this paper, a lane-level RM mapping system using a monocular camera is developed. The developed system can be considered as an alternative to expensive high-end MMS. The developed RM map includes the information of road lanes (RLs) and symbolic road markings (SRMs). First, to build a lane-level RM map, the RMs are segmented at pixel level through the deep learning network. The network is named RMNet. The segmented RMs are then gathered to build a lane-level RM map. Second, the lane-level map is improved through loop-closure detection and graph optimization. To train the RMNet and build a lane-level RM map, a new dataset named SeRM set is developed. The set is a large dataset for lane-level RM mapping and it includes a total of 25157 pixel-wise annotated images and 21000 position labeled images. Finally, the proposed lane-level map building method is applied to SeRM set and its validity is demonstrated through experimentation.

A Hybrid Ensemble Model for Urban Lane-Level Traffic Flow Prediction

In recent years, automatic driving and internet of vehicles technology have made great progress. Road oriented the traditional traffic flow prediction methods can no longer meet the needs of automatic driving, lane-level map navigation. This paper takes urban lanes as the research object, and proposes a traffic flow prediction model for urban lanes. First, a boosting based Catboost model is introduced to construct a series of spatiotemporal features and perform feature selection to reduce the model bias. Second, the model variance is reduced by using bagging-based random forest algorithm. Third, a long short-term memory network (LSTM) is used to extract the temporal trend of traffic flow in the current lane. The prediction results of these three models are finally integrated by the method of stacking ensemble learning. The proposed method is evaluated with the dataset collected from real intersections in Wuxi city. The experimental results show that our model has higher prediction precision than other methods.

Ego-Lane Index Estimation Based on Lane-Level Map and LiDAR Road Boundary Detection.

Correct ego-lane index estimation is essential for lane change and decision making for intelligent vehicles, especially in global navigation satellite system (GNSS)-challenged environments. To achieve this, we propose an ego-lane index estimation approach in an urban scenario based on particle filter (PF). The particles are initialized and propagated by dead reckoning with inertial measurement unit (IMU) and odometry. A lane-level map is used to navigate the particles taking advantage of topologic and geometric information of lanes. GNSS single-point positioning (SPP) can provide position estimation with meter-level accuracy in urban environments, which can limit drift introduced by dead reckoning for updating the weight of each particle. Light detection and ranging (LiDAR) is a common sensor in an intelligent vehicle. A LiDAR-based road boundary detection method provides distance measurements from the vehicle to the left/right road boundaries, which provides a measurement for importance weighting. However, the high precision of the LiDAR measurements may put a tight constraint on the distribution of particles, which can lead to particle degeneration with sparse particle sets. To deal with this problem, we propose a novel step that shifts particles laterally based on LiDAR measurements instead of importance weighting in the traditional PF scheme. We tested our methods on an urban expressway at a low traffic volume period to ensure road boundaries can be detected by LiDAR measurements at most time steps. Experimental results prove that our improved PF scheme can correctly estimate ego-lane index at all time steps, while the traditional PF scheme produces wrong estimations at some time steps.

Lane-Level Map-Aiding Approach Based on Non-Lane-Level Digital Map Data in Road Transport Security

To prevent terror attacks in which trucks are used as weapons as happened in Nice or Berlin in 2016, the European Project Autonomous Emergency Maneuvering and Movement Monitoring for Road Transport Security (TransSec) was launched in 2018. One crucial point of this project is the development of a map-aiding approach for the localization of vehicles on digital maps, so that the information in digital map data can be used to detect prohibited driving maneuvers, such as off-road or wrong-way drivers. For example, a lane-level map-aiding approach is required for wrong-way driver detection. Navigation Data Standard (NDS) is one of the worldwide map standards developed by several automobile manufacturers. So far, there is no lane-level NDS map covers a large area, therefore, it was decided to use the latest available NDS map without lane level accuracy. In this paper, a lane-level map-aiding approach based on a non-lane-level NDS map is presented. Due to the inaccuracy of vehicle position and digital map the map-aiding does not always provide the correct results, so probabilities of off-road and wrong-way diver detection are estimated to support risk estimation. The performance of the developed map-aiding approach is comprehensively evaluated with both real and simulated trajectories.

Efficient Lane-Level Map Building via Vehicle-Based Crowdsourcing

By providing rich context of lane information on roads, lane-level maps play a vital role in intelligent transportation systems. Since Global Positioning Systems (GPS) have been widely applied to vehicles, vehicle-based crowdsourcing offers an economical way to the lane-level map building by collecting and analyzing the GPS trajectories of vehicles. However, existing works cannot directly extract lane-level road information from raw and interleaved crowdsourcing trajectories, and moreover they are time-consuming and inaccurate. In this article, we propose a lane-level map building scheme, which can directly extract lane-level road information from raw crowdsourcing GPS trajectories with both efficiency and accuracy improvement. Consider the global similarity between trajectories, we design an efficient trajectory segmentation and clustering algorithm based on improved discrete Fr&#x00E9;chet distance and entropy theory, which can directly and accurately deal with the interleaved and messy trajectories. To improve the efficiency, we employ the Least Square Estimate (LSE) to constrain Gaussian Mixture Model (GMM) and design an efficient and accurate lane-level road information extraction algorithm. Comprehensive comparative experiments and performance evaluation on a real-world trajectory dataset show that the proposed scheme outperforms the state-of-the-art works in terms of both efficiency and accuracy.

A Lightweight Simulation Framework for Learning Control Policies for Autonomous Vehicles in Real-World Traffic Condition

We present a new simulation framework for learning control policies for autonomous vehicles (AVs) based on real-world vehicle data and maps. The framework we propose consists of three major components: a) creating a detailed lane-level map (i.e., a high definition (HD) map) for a region of interest, b) generating an environment on the HD-map using the sensor outputs (e.g., GPS, radar) from vehicles driving in the same region, and c) learning a control policy based on the realistic environment constructed. We created the lane-level HD-maps using open street maps (OSM) and aerial imagery, from which we extracted the lane-level marking and edge features. The extracted image features are then utilized to calculate higher level attributes (e.g., curvature, heading, cross-sections etc.) for each point in the HD-map. The data acquired from vehicle sensors is combined with the constructed map to create a realistic environment. Based on the constructed environment, we learned a policy to control the vehicle laterally using a reinforcement learning algorithm and longitudinally using proportional-integral-derivative (PID) controller. Our experimental results show that the proposed framework works well, offering a flexible and scalable solution for learning control policies for AVs in realistic environments.

Lane-Level Map Matching Based on HMM

Lane-level map matching is essential for autonomous driving. In this paper, we propose a Hidden Markov Model (HMM) for matching a trajectory of noisy GPS measurements to the road lanes in which the vehicle records its positions. To our knowledge, this is the first time that HMM is used for lane-level map matching. Apart from GPS values, the model is further assisted by yaw rate data (converted to a lane change indicator signal) and visual cues in the form of the left and right lane marking types (dashed, solid, etc.). Having defined expressions for the HMM emission and transition probabilities, we evaluate our model to demonstrate that it achieves 95.1% recall and 3.3% median path length error for motorway trajectories.

Precise Single-Frequency Positioning Using Low-Cost Receiver with the Aid of Lane-Level Map Matching for Land Vehicle Navigation

Precise positioning with low-cost single-frequency global navigation satellite system (GNSS) receivers has great potential in a wide range of applications because of its low price and improved accuracy. However, challenges remain in achieving reliable and accurate solutions using low-cost receivers. For instance, the successful ambiguity fixing rate could be low for real-time kinematic (RTK) while large errors may occur in precise point positioning (PPP) in some scenarios (e.g., trees along the road). To solve the problems, this paper proposes a method with the aid of additional lane-level digital map information to improve the accuracy and reliability of RTK and PPP solutions. In the method, a digital camera will be applied for lane recognition and the positioning solution from a low-cost receiver will be projected to the digital map lane link. With the projected point position as a constraint, the RTK ambiguity fixing rate and PPP performance can be enhanced. A field kinematic test was conducted to verify the improvement of the RTK and PPP solutions with the aid of map matching. The results show that the RTK ambiguity fixing rate can be increased and the PPP positioning error can be reduced by map matching.

Automatically Building Linking Relations between Lane-Level Map and Commercial Navigation Map Using Topological Networks Matching

The lane-level map, which contains the lane-level information severely lacking in widely used commercial navigation maps, has become an essential data source for autonomous driving systems. The linking relations between lane-level map and commercial navigation map can facilitate an autonomous driving system mapping information between different applications using different maps. In this paper, an approach is proposed to build the linking relations automatically. The different topology networks are first reconstructed into similar structures. Then, to build the linking relations automatically, the adaptive multi-filter algorithm and forward path exploring algorithm are proposed to detect corresponding junctions and paths, respectively. The approach is validated by two real data sets of more than 150 km of roads, mainly highway. The linking relations for nearly 94% of the total road length have been built successfully.

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

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.

Generation of Lane-Level Road Networks Based on a Trajectory-Similarity-Join Pruning Strategy

With the development of autonomous driving, lane-level maps have attracted significant attention. Since the lane-level road network is an important part of the lane-level map, the efficient, low-cost, and automatic generation of lane-level road networks has become increasingly important. We propose a new method here that generates lane-level road networks using only position information based on an autonomous vehicle and the existing lane-level road networks from the existing road-level professionally surveyed without lane details. This method uses the parallel relationship between the centerline of a lane and the centerline of the corresponding segment. Since the direct point-by-point computation is huge, we propose a method based on a trajectory-similarity-join pruning strategy (TSJ-PS). This method uses a filter-and-verify search framework. First, it performs quick segmentation based on the minimum distance and then uses the similarity of two trajectories to prune the trajectory similarity join. Next, it calculates the centerline trajectory for lanes using the simulation transformation model by the unpruned trajectory points. Finally, we demonstrate the efficiency of the algorithm and generate a lane-level road network via experiments on a real road.

Lane-Level Road Network Generation Techniques for Lane-Level Maps of Autonomous Vehicles: A Survey

Autonomous driving is experiencing rapid development. A lane-level map is essential for autonomous driving, and a lane-level road network is a fundamental part of a lane-level map. A large amount of research has been performed on lane-level road network generation based on various on-board systems. However, there is a lack of analysis and summaries with regards to previous work. This paper presents an overview of lane-level road network generation techniques for the lane-level maps of autonomous vehicles with on-board systems, including the representation and generation of lane-level road networks. First, sensors for lane-level road network data collection are discussed. Then, an overview of the lane-level road geometry extraction methods and mathematical modeling of a lane-level road network is presented. The methodologies, advantages, limitations, and summaries of the two parts are analyzed individually. Next, the classic logic formats of a lane-level road network are discussed. Finally, the survey summarizes the results of the review.

A Flexible Multi-Layer Map Model Designed for Lane-Level Route Planning in Autonomous Vehicles

Abstract An increasing number of drivers are relying on digital map navigation systems in vehicles or mobile phones to select optimal driving routes in order to save time and improve safety. In the near future, digital map navigation systems are expected to play more important roles in transportation systems. In order to extend current navigation systems to more applications, two fundamental problems must be resolved: the lane-level map model and lane-level route planning. This study proposes solutions to both problems. The current limitation of the lane-level map model is not its accuracy but its flexibility; this study proposes a novel seven-layer map structure, called as Tsinghua map model, which is able to support autonomous driving in a flexible and efficient way. For lane-level route planning, we propose a hierarchical route-searching algorithm to accelerate the planning process, even in the presence of complicated lane networks. In addition, we model the travel costs allocated for lane-level road networks by analyzing vehicle maneuvers in traversing lanes, changing lanes, and turning at intersections. Tests were performed on both a grid network and a real lane-level road network to demonstrate the validity and efficiency of the proposed algorithm.

Real-Time Road Lane Detection in Urban Areas Using LiDAR Data

The generation of digital maps with lane-level resolution is rapidly becoming a necessity, as semi- or fully-autonomous driving vehicles are now commercially available. In this paper, we present a practical real-time working prototype for road lane detection using LiDAR data, which can be further extended to automatic lane-level map generation. Conventional lane detection methods are limited to simple road conditions and are not suitable for complex urban roads with various road signs on the ground. Given a 3D point cloud scanned by a 3D LiDAR sensor, we categorized the points of the drivable region and distinguished the points of the road signs on the ground. Then, we developed an expectation-maximization method to detect parallel lines and update the 3D line parameters in real time, as the probe vehicle equipped with the LiDAR sensor moved forward. The detected and recorded line parameters were integrated to build a lane-level digital map with the help of a GPS/INS sensor. The proposed system was tested to generate accurate lane-level maps of two complex urban routes. The experimental results showed that the proposed system was fast and practical in terms of effectively detecting road lines and generating lane-level maps.