Efficient edge-aware surface mesh reconstruction for urban scenes

Bdis-Szomor, Andrs; Riemenschneider, Hayko; Gool, Luc Van

doi:10.1016/j.cviu.2016.06.002

Cited by 27 publications

(12 citation statements)

References 82 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…No SLAM techniques can currently build higher-level representations, beyond point clouds, mesh models, surfels models, and TSDFs. Recent efforts in this direction include [17,51,231].…”

Section: Representation I: Metric Map Modelsmentioning

confidence: 99%

Past, Present, and Future of Simultaneous Localization and Mapping: Toward the Robust-Perception Age

Cadena¹,

Carlone²,

Carrillo³

et al. 2016

IEEE Trans. Robot.

3,141

1,742

View full text Add to dashboard Cite

Abstract-Simultaneous localization and mapping (SLAM) consists in the concurrent construction of a model of the environment (the map), and the estimation of the state of the robot moving within it. The SLAM community has made astonishing progress over the last 30 years, enabling large-scale real-world applications and witnessing a steady transition of this technology to industry. We survey the current state of SLAM and consider future directions. We start by presenting what is now the de-facto standard formulation for SLAM. We then review related work, covering a broad set of topics including robustness and scalability in long-term mapping, metric and semantic representations for mapping, theoretical performance guarantees, active SLAM and exploration, and other new frontiers. This paper simultaneously serves as a position paper and tutorial to those who are users of SLAM. By looking at the published research with a critical eye, we delineate open challenges and new research issues, that still deserve careful scientific investigation. The paper also contains the authors' take on two questions that often animate discussions during robotics conferences: Do robots need SLAM? and Is SLAM solved?

show abstract

“…No SLAM techniques can currently build higher-level representations, beyond point clouds, mesh models, surfels models, and TSDFs. Recent efforts in this direction include [17,51,231].…”

Section: Representation I: Metric Map Modelsmentioning

confidence: 99%

Past, Present, and Future of Simultaneous Localization and Mapping: Toward the Robust-Perception Age

Cadena¹,

Carlone²,

Carrillo³

et al. 2016

IEEE Trans. Robot.

3,141

1,742

View full text Add to dashboard Cite

show abstract

“…Other sparse methods ( [27,39,37,13,18,2]) use 2D (not 3D) Delaunay triangulations constrained in images by interest points and image contours. However, [27,39] are limited to simple topology (plane/sphere), large scale scenes cannot be computed efficiently by [37] due to the use of regular decomposition of the 3D space, and [13,18,2] do not generate a globally consistent manifold surface but several (view-centered) local models with redundancy. There are also methods that need strong assumptions to segment the scene using planar regions ( [41,9]) or swept surfaces for architectural scenes ( [43]).…”

Section: Previous Sparse Methodsmentioning

confidence: 99%

Surface reconstruction from a sparse point cloud by enforcing visibility consistency and topology constraints

Lhuillier

2018

Computer Vision and Image Understanding

View full text Add to dashboard Cite

Highlights• Surface reconstruction under-uses topology constraints in Computer Vision• Start from a previous method enforcing manifoldness on a sparse point cloud• Simultaneously enforce visibility consistency and low genus for the first time• Improve the removal of surface singularities and the escape from local extrema• Experiment on long video sequences taken by a helmetheld omnidirectional camera AbstractThere are reasons to reconstruct a surface from a sparse cloud of 3D points estimated from an image sequence: to avoid computationally expensive dense stereo, e.g. for applications that do not need high level of details and have limited resources, or to initialize dense stereo in other cases. It is also interesting to enforce topology constraints (like manifoldness) for both surface regularization and applications. In this article, we improve by several ways a previous method that enforces the manifold constraint given a sparse point cloud. We enforce lowered genus, i.e. simplified topology, as a further regularization constraint for maximizing the visibility consistency encoded in a 3D Delaunay triangulation of the points. We also provide more efficient escapes from local extrema, an acceleration of the manifold test and more efficient removals of surface singularities. We experiment on a sparse point cloud reconstructed from videos, that are taken by a helmet-held omnidirectional multi-camera moving in an university campus.

show abstract

“…As seen in Riemenschneider, H., et al and Bódis-Szomorú et al, the respective approaches have motivated us in the sense that façade 2D information can be explored more thoroughly in order to improve its volumetry reconstruction [31,32]. Martinovic, A., et al, for instance, used the extracted façade features to analyze the alignment among them, where a simple discontinuity showed the boundaries between different façades [46].…”

Section: Advances In 3d Urban Reconstructionmentioning

confidence: 99%

“…That means not only a good method would be required, but the addition of another variable, such as geometry, should be used to improve class separability. A new demand in the areas of photogrammetry and remote sensing is leading the research to further analysis of these urban objects, taking advantage of the aforementioned optical campaigns, such as in [31][32][33][34], to acquire the object geometries in a low-cost and simple-to-use manner.…”

Section: Introductionmentioning

confidence: 99%

3D Façade Labeling over Complex Scenarios: A Case Study Using Convolutional Neural Network and Structure-From-Motion

et al. 2018

View full text Add to dashboard Cite

Urban environments are regions in which spectral variability and spatial variability are extremely high, with a huge range of shapes and sizes, and they also demand high resolution images for applications involving their study. Due to the fact that these environments can grow even more over time, applications related to their monitoring tend to turn to autonomous intelligent systems, which together with remote sensing data could help or even predict daily life situations. The task of mapping cities by autonomous operators was usually carried out by aerial optical images due to its scale and resolution; however new scientific questions have arisen, and this has led research into a new era of highly-detailed data extraction. For many years, using artificial neural models to solve complex problems such as automatic image classification was commonplace, owing much of their popularity to their ability to adapt to complex situations without needing human intervention. In spite of that, their popularity declined in the mid-2000s, mostly due to the complex and time-consuming nature of their methods and workflows. However, newer neural network architectures have brought back the interest in their application for autonomous classifiers, especially for image classification purposes. Convolutional Neural Networks (CNN) have been a trend for pixel-wise image segmentation, showing flexibility when detecting and classifying any kind of object, even in situations where humans failed to perceive differences, such as in city scenarios. In this paper, we aim to explore and experiment with state-of-the-art technologies to semantically label 3D urban models over complex scenarios. To achieve these goals, we split the problem into two main processing lines: first, how to correctly label the façade features in the 2D domain, where a supervised CNN is used to segment ground-based façade images into six feature classes, roof, window, wall, door, balcony and shop; second, a Structure-from-Motion (SfM) and Multi-View-Stereo (MVS) workflow is used to extract the geometry of the façade, wherein the segmented images in the previous stage are then used to label the generated mesh by a “reverse” ray-tracing technique. This paper demonstrates that the proposed methodology is robust in complex scenarios. The façade feature inferences have reached up to 93% accuracy over most of the datasets used. Although it still presents some deficiencies in unknown architectural styles and needs some improvements to be made regarding 3D-labeling, we present a consistent and simple methodology to handle the problem.

show abstract

Efficient edge-aware surface mesh reconstruction for urban scenes

Cited by 27 publications

References 82 publications

Past, Present, and Future of Simultaneous Localization and Mapping: Toward the Robust-Perception Age

Past, Present, and Future of Simultaneous Localization and Mapping: Toward the Robust-Perception Age

Surface reconstruction from a sparse point cloud by enforcing visibility consistency and topology constraints

3D Façade Labeling over Complex Scenarios: A Case Study Using Convolutional Neural Network and Structure-From-Motion

Contact Info

Product

Resources

About