Experiments to Distribute and Parallelize Map Generalization Processes

Touya, Guillaume; Berli, Justin; Lokhat, Imran; Regnauld, Nicolas

doi:10.1080/00087041.2017.1413787

Cited by 10 publications

(13 citation statements)

References 14 publications

(17 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Experiments were carried out to parallelize the algorithms from CartAGen (Touya et al, 2017b), to show that they can be applied to very large datasets, using a third party framework.…”

Section: Single Objectsmentioning

confidence: 99%

CartAGen: an Open Source Research Platform for Map Generalization

Touya¹,

Lokhat²,

Duchêne³

2019

Proc. Int. Cartogr. Assoc.

Self Cite

View full text Add to dashboard Cite

<p><strong>Abstract.</strong> Automatic map generalization is a complex task that is still a research problem and requires the development of research prototypes before being usable in productive map processes. In the meantime, reproducible research principles are becoming a standard. Publishing reproducible research means that researchers share their code and their data so that other researchers might be able to reproduce the published experiments, in order to check them, extend them, or compare them to their own experiments. Open source software is a key tool to share code and software, and CartAGen is the first open source research platform that tackles the overall map generalization problem: not only the building blocks that are generalization algorithms, but also methods to chain them, and spatial analysis tools necessary for data enrichment. This paper presents the CartAGen platform, its architecture and its components. The main component of the platform is the implementation of several multi-agent based models of the literature such as AGENT, CartACom, GAEL, CollaGen, or DIOGEN. The paper also explains and discusses different ways, as a researcher, to use or to contribute to CartAGen.</p>

show abstract

“…Experiments were carried out to parallelize the algorithms from CartAGen (Touya et al, 2017b), to show that they can be applied to very large datasets, using a third party framework.…”

Section: Single Objectsmentioning

confidence: 99%

CartAGen: an Open Source Research Platform for Map Generalization

Touya¹,

Lokhat²,

Duchêne³

2019

Proc. Int. Cartogr. Assoc.

Self Cite

View full text Add to dashboard Cite

show abstract

“…Therefore, a partitioning method based on a quadtree grid was developed by Briat, Monnot, and Punt (2011) and Thiemann, Werder, Globig, and Sester (2013) to achieve the balanced utilization of computing resources by ensuring that the data in each partitioning cell are basically the same. However, the problem of grid boundaries crossing buildings is prone to occur in the above two methods because a standard regular grid cannot delineate the spatial distribution of buildings (Touya, Berli, Lokhat, & Regnauld, 2017). To address this issue, scholars have attempted to perform building partitioning based on geographic units, such as administrative regions, watersheds, and road networks.…”

Section: Introductionmentioning

confidence: 99%

“…To address this issue, scholars have attempted to perform building partitioning based on geographic units, such as administrative regions, watersheds, and road networks. Because road networks can provide essential spatial constraints in the building generalization process, partitioning based on road networks has been widely accepted and adopted (Regnauld, Touya, Gould, & Foerster, 2014; Touya et al, 2017; Wang & Doihara, 2004). Specifically, large‐scale buildings are partitioned with small‐scale road networks (as shown in Figure 3), and the buildings in each partitioning cell are processed separately and in parallel.…”

Section: Introductionmentioning

confidence: 99%

A multi‐scale partitioning and aggregation method for large volumes of buildings considering road networks association constraints

Yin

et al. 2021

Transactions in GIS

View full text Add to dashboard Cite

The automatic aggregation of buildings often requires the handling of large volumes of data, which results in high processing overhead. One efficient and feasible way of managing this issue is to partition data using a road network and perform parallel computing. Generally, road networks are used not only to divide buildings into multiple cells but also to provide strong spatial constraints for the geometric morphology and a balanced spatial distribution of building aggregation results. Existing methods have focused on the former at the expense of the latter, which easily leads to spatial conflicts between road networks, and building aggregation results in the multi‐scale representation of buildings. Hence, a multi‐scale partitioning and aggregation method for large volumes of buildings considering the association constraint of the road network is proposed. First, road stroke connections are established, and road levels are calculated on the basis of semantic, geometric, and topological features of road strokes. Second, large volumes of buildings are partitioned level‐by‐level based on multi‐level roads and the convex hull of buildings. Finally, a “three morphological transformation algorithm” is proposed to aggregate buildings in each partitioning cell. OpenStreetMap data (31 × 33 km2) from Tokyo Bay, Japan, were used for validation. The experiment reveals that compared to the state‐of‐the‐art method, the proposed method avoids spatial crossing conflicts and eliminates “narrow necks” in the building aggregation results. The coordination between buildings and the road network was enhanced, and the morphological similarity of the buildings before and after aggregation was improved.

show abstract

“…The problems raised by both methods are similar to classical problems in data partitioning for map generalisation [33], where there is no perfect solution. There are two important issues to consider here: first the scale of the images should be fixed to help the convolutional network to learn that scale changes from the input image to the output image.…”

mentioning

confidence: 98%

Exploring the Potential of Deep Learning Segmentation for Mountain Roads Generalisation

Courtial

Ayedi

Touya

et al. 2020

IJGI

Self Cite

View full text Add to dashboard Cite

Among cartographic generalisation problems, the generalisation of sinuous bends in mountain roads has always been a popular one due to its difficulty. Recent research showed the potential of deep learning techniques to overcome some remaining research problems regarding the automation of cartographic generalisation. This paper explores this potential on the popular mountain road generalisation problem, which requires smoothing the road, enlarging the bend summits, and schematising the bend series by removing some of the bends. We modelled the mountain road generalisation as a deep learning problem by generating an image from input vector road data, and tried to generate it as an output of the model a new image of the generalised roads. Similarly to previous studies on building generalisation, we used a U-Net architecture to generate the generalised image from the ungeneralised image. The deep learning model was trained and evaluated on a dataset composed of roads in the Alps extracted from IGN (the French national mapping agency) maps at 1:250,000 (output) and 1:25,000 (input) scale. The results are encouraging as the output image looks like a generalised version of the roads and the accuracy of pixel segmentation is around 65%. The model learns how to smooth the output roads, and that it needs to displace and enlarge symbols but does not always correctly achieve these operations. This article shows the ability of deep learning to understand and manage the geographic information for generalisation, but also highlights challenges to come.

show abstract

Experiments to Distribute and Parallelize Map Generalization Processes

Cited by 10 publications

References 14 publications

CartAGen: an Open Source Research Platform for Map Generalization

CartAGen: an Open Source Research Platform for Map Generalization

A multi‐scale partitioning and aggregation method for large volumes of buildings considering road networks association constraints

Exploring the Potential of Deep Learning Segmentation for Mountain Roads Generalisation

Contact Info

Product

Resources

About