Automatic Vector-Based Road Structure Mapping Using Multibeam LiDAR

Zhao, Junqiao; He, Xueming; Li, Jun; Feng, Tiantian; Ye, Chen; Liu, Xiong

doi:10.3390/rs11141726

Cited by 15 publications

(4 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Iesaki et al [13] presented a method for generating polynomial curves between intersection road lanes, fitted according to a learned cost function. Zhao et al [14] presented a SLAMbased method to generate a closed vector map including road lanes. Guo et al [15] presented a method for generating a lane-level road network graph based on superimposed vehicle trajectories and road markings.…”

Section: A Automatic and Semi-automatic Lane-level Map Generationmentioning

confidence: 99%

“…After reversing all path directions, the same unification algorithm can be applied to unify the reverse tree from the end to the point where all paths converge into a single path, returning the common exit path p exit and a new set of paths P * inter forming a reverse tree into the start of p exit (line 13). These paths are added to the maintained sets (lines [14][15]. Repeating these steps for all exit points results in the set P exit constituting the paths of all exit lanes in the road scene.…”

Section: B Search-based Path Algorithmmentioning

confidence: 99%

See 1 more Smart Citation

Learning a Model for Inferring a Spatial Road Lane Network Graph using Self-Supervision

Karlsson¹,

Wong²,

Thompson³

et al. 2021

Preprint

View full text Add to dashboard Cite

Interconnected road lanes are a central concept for navigating urban roads. Currently, most autonomous vehicles rely on preconstructed lane maps as designing an algorithmic model is difficult. However, the generation and maintenance of such maps is costly and hinders large-scale adoption of autonomous vehicle technology. This paper presents the first self-supervised learning method to train a model to infer a spatially grounded lane-level road network graph based on a dense segmented representation of the road scene generated from onboard sensors. A formal road lane network model is presented and proves that any structured road scene can be represented by a directed acyclic graph of at most depth three while retaining the notion of intersection regions, and that this is the most compressed representation. The formal model is implemented by a hybrid neural and search-based model, utilizing a novel barrier function loss formulation for robust learning from partial labels. Experiments are conducted for all common road intersection layouts. Results show that the model can generalize to new road layouts, unlike previous approaches, demonstrating its potential for real-world application as a practical learning-based lane-level map generator.

show abstract

Section: A Automatic and Semi-automatic Lane-level Map Generationmentioning

confidence: 99%

Section: B Search-based Path Algorithmmentioning

confidence: 99%

Learning a Model for Inferring a Spatial Road Lane Network Graph using Self-Supervision

Karlsson¹,

Wong²,

Thompson³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Additionally, the model relies heavily on road markings and is thus heavily dependent on a particular feature. Zhao et al [19] used human driving data with a SLAM-based approach to generate a vectorized road lane map. These semiautomatic HD map generation methods do not generalize by learning a model of the road network through contextual features, and thus cannot be applied to new environments unlike our proposed method.…”

Section: B Semi-automatic Hd Map Generationmentioning

confidence: 99%

Learning a Directional Soft Lane Affordance Model for Road Scenes Using Self-Supervision

Karlsson¹,

Sjoberg²

2020

Preprint

View full text Add to dashboard Cite

Humans navigate complex environments in an organized yet flexible manner, adapting to the context and implicit social rules. Understanding these naturally learned patterns of behavior is essential for applications such as autonomous vehicles. However, algorithmically defining these implicit rules of human behavior remains difficult. This work proposes a novel self-supervised method for training a probabilistic network model to estimate the regions humans are most likely to drive in as well as a multimodal representation of the inferred direction of travel at each point. The model is trained on individual human trajectories conditioned on a representation of the driving environment. The model is shown to successfully generalize to new road scenes, demonstrating potential for realworld application as a prior for socially acceptable driving behavior in challenging or ambiguous scenarios which are poorly handled by explicit traffic rules.

show abstract

“…For building extraction, this problem is more critical since the background and scenarios of the VHR remote sensing image are much more complex and diverse, and the shape of the building is tremendously more regular and sharper than that of the natural objects. Blur and inaccurate boundaries seriously affect the quality of visual evaluation and further building vectorization [13]. To overcome this problem, within the semantic information obtained with a deep CNN model, some researchers have attempted to fuse multisource images such as lidar images, SAR images, and DEM images.…”

Section: Introductionmentioning

confidence: 99%

An Improved Boundary-Aware Perceptual Loss for Building Extraction from VHR Images

Zhang

Gong

et al. 2020

Remote Sensing

View full text Add to dashboard Cite

With the development of deep learning technology, an enormous number of convolutional neural network (CNN) models have been proposed to address the challenging building extraction task from very high-resolution (VHR) remote sensing images. However, searching for better CNN architectures is time-consuming, and the robustness of a new CNN model cannot be guaranteed. In this paper, an improved boundary-aware perceptual (BP) loss is proposed to enhance the building extraction ability of CNN models. The proposed BP loss consists of a loss network and transfer loss functions. The usage of the boundary-aware perceptual loss has two stages. In the training stage, the loss network learns the structural information from circularly transferring between the building mask and the corresponding building boundary. In the refining stage, the learned structural information is embedded into the building extraction models via the transfer loss functions without additional parameters or postprocessing. We verify the effectiveness and efficiency of the proposed BP loss both on the challenging WHU aerial dataset and the INRIA dataset. Substantial performance improvements are observed within two representative CNN architectures: PSPNet and UNet, which are widely used on pixel-wise labelling tasks. With BP loss, UNet with ResNet101 achieves 90.78% and 76.62% on IoU (intersection over union) scores on the WHU aerial dataset and the INRIA dataset, respectively, which are 1.47% and 1.04% higher than those simply trained with the cross-entropy loss function. Additionally, similar improvements (0.64% on the WHU aerial dataset and 1.69% on the INRIA dataset) are also observed on PSPNet, which strongly supports the robustness of the proposed BP loss.

show abstract

Automatic Vector-Based Road Structure Mapping Using Multibeam LiDAR

Cited by 15 publications

References 45 publications

Learning a Model for Inferring a Spatial Road Lane Network Graph using Self-Supervision

Learning a Model for Inferring a Spatial Road Lane Network Graph using Self-Supervision

Learning a Directional Soft Lane Affordance Model for Road Scenes Using Self-Supervision

An Improved Boundary-Aware Perceptual Loss for Building Extraction from VHR Images

Contact Info

Product

Resources

About