Feed on
Posts
Comments

A new paper on tree height estimation from TanDEM-X data has just been published in Remote Sensing of Environment. The article finds that tree height can be predicted using TanDEM-X metrics (backscatter, bistatic coherence, and interferometric height) in the sparse forest patches of the Arctic treeline zone at the transition from forest to tundra. Taking into account the global coverage of bistatic TanDEM-X data acquired for the global digital elevation model, the paper shows the potential for quantifying the tree height in small forest patches along the entire circum-Arctic treeline zone.

Antonova, S., Thiel, C., Höfle, B., Anders, K., Helm, V., Zwieback, S., Marx, S., Boike, J. (2019). Estimating tree height from TanDEM-X data at the northwestern Canadian treeline. Remote Sensing of Environment, 231,  doi: 10.1016/j.rse.2019.111251.

Use this link to get free access to the paper until 06th August 2019.

Want to have a look at tree height and other surface properties in the study region? The airborne LiDAR data is published and openly accessible in the PANGAEA data library.

The research was conducted within the PermaSAR project by the Permafrost Research Unit and the 3DGeo Research Group and funded by the BMWi/DLR in the framework “Entwicklung von innovativen wissenschaftlichen Methoden und Produkten im Rahmen der TanDEM-X Science Phase”.

This week, the 3DGeo participated in the ISPRS Geospatial Week 2019 with two presentations among the sessions of the Laser Scanning Workshop with many interesting talks and poster.

Presentations were given by Ashutosh Kumar in the Machine Learning Session and Katharina Anders in the Change Detection Session.

Highlight: The work by Ashutosh Kumar on feature relevance in deep learning for 3D point cloud classification got awarded the prize for best paper!

The work in this paper was performed during Ashutosh Kumars internship in the 3DGeo in summer 2018 with the successful conclusion in a paper published in the ISPRS Annals of Photogrammetry, Remote Sensing and Spatial Information Science. We thank the Organizing and Scientific Committees of the ISPRS Workshop Laser Scanning for their constructive review and appreciation of this paper!

The talks are over, but some impressions were captured…

… tweeted about: #3DGeo #ISPRSgsw

… and the full papers are openly accessible:

Anders, K., Lindenbergh, R. C., Vos, S. E., Mara, H., de Vries, S., and Höfle, B. (2019). HIGH-FREQUENCY 3D GEOMORPHIC OBSERVATION USING HOURLY TERRESTRIAL LASER SCANNING DATA OF A SANDY BEACH, ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci., IV-2/W5, 317-324, DOI: 10.5194/isprs-annals-IV-2-W5-317-2019

Kumar, A., Anders, K., Winiwarter, L., and Höfle, B. (2019). FEATURE RELEVANCE ANALYSIS FOR 3D POINT CLOUD CLASSIFICATION USING DEEP LEARNING, ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci., IV-2/W5, 373-380, DOI: 10.5194/isprs-annals-IV-2-W5-373-2019

Geospatial data is key for empowering citizens around the globe and to achieve the SDGs— if geodata is made openly available and easy to be put to use. The Humanitarian OpenStreetMap Team (HOT) is in this regard coordinating and supporting humanitarian action and community resilience through open mapping.

The GIScience Research Group has supported HOT’s work for a couple of years already, through research and application development, to support open map data creation, analyses, and use. These efforts have shown great potential for humanitarian projects and disaster management and are since 2016 also further supported and strengthened by the Heidelberg Institute for Geoinformation Technology (HeiGIT).

The May/ June edition of the Geospatial World Magazine covers some of these activities in their print version, see the article here, pages 54-56. The article is a follow-up on a related presentation by Melanie Eckle (HeiGIT/HOT) during the Geo4SDGs session at the Geospatial World Forum that was organized in April in Amsterdam.

An extension of the article will be published beginning of July in the online version of the magazine- so stay tuned.

The Geospatial Industry Magazine

Latest issue of Geospatial World: The Geospatial Industry Magazine

Land is a spare resource so it makes sense to think about how to use it most efficiently. This leads to the problem of land use allocation under consideration of trade-offs. Multi-objective optimization algorithms are a tool quantify the trade-offs by estimating the Pareto-optimal land use allocations. Often, constraints in the solution space have to be incorporated in land use optimization approaches: most frequently, the amount of land use change allowed or transitions between land use classes is restricted. A group of scientist - involving GIScience/HeiGIT member - Sven Lautenbach has published a python library for constraint handling in land use optimization. Two options have been implemented for constraint handling: penalization or use of repair mechanism. Testing both options on virtual test data showed that the repair mechanism for constraint handling outperformed the approach based on penalization. The test data represent a virtual landscape, the control variables represent land use at patch level and the objective function consists of crop yield, forest species richness, habitat heterogeneity and water yield.

Non-dominated solutions of a constrained patch-level three-objective
optimization run using repair mutation considering both transition and area
rules compared to the status quo land use with illustration of extreme and mid-
range solutions. Max CY is the solution with maximal crop yield, Max SR the
solution with maximal species richness and Max HH the solution with maximal
habitat heterogeneity. Values were normalized by the theoretical (un-
constrained) optimum values for each objective.
Strauch, Michael, Anna F. Cord, Carola Pätzold, Sven Lautenbach, Andrea Kaim, Christian Schweitzer, Ralf Seppelt, und Martin Volk. „Constraints in Multi-Objective Optimization of Land Use Allocation – Repair or Penalize?“ Environmental Modelling & Software, Mai 2019. https://doi.org/10.1016/j.envsoft.2019.05.003.
The article can be freely accessed in the next 35 days.

Namibia is a dry and low populated country highly dependent on agriculture, with many areas experiencing land degradation accelerated by climate change. One of the most obvious and damaging manifestations of these degradation processes are gullies, which lead to great economic losses while accelerating desertification.

The development of standardized methods to detect and monitor the evolution of gully-affected areas is crucial to plan prevention and remediation strategies.

This paper explores fully automated satellite-based remote sensing methods with the aim of developing solutions applicable at a regional or even national scale. For this purpose, three different algorithms are applied to a Digital Elevation Model (DEM) generated from the TanDEM-X satellite mission to extract gullies from their geomorphological characteristics: (i) Inverted Morphological Reconstruction (IMR), (ii) Smoothing Moving Polynomial Fitting (SMPF) and (iii) Multi Profile Curvature Analysis (MPCA). These algorithms are adapted or newly developed to identify gullies at the pixel level (12 m) in our study site in the Krumhuk Farm. The results of the three methods are benchmarked with ground truth; specific scenarios are observed to better understand the performance of each method.

XX

Results of the different gully classification methods for one validation plot.

Results show that MPCA is the most reliable method to identify gullies, achieving an overall accuracy of ca. 0.80 with values of Cohen Kappa close to 0.35. The performance of these parameters improves when detecting large gullies (>30 m width and >3 m depth) achieving Total Accuracies (TA) near to 0.90, Cohen Kappa above 0.5, and User Accuracy (UA) and Producer Accuracy (PA) over 0.50 for the gully class. Small gullies (<12 m wide and <2 m deep) are usually neglected in the classification results due to spatial resolution constraints within the input DEM. In addition, IMR generates accurate results for UA in the gully class (0.94).

The MPCA method developed here is a promising tool for the identification of large gullies considering extensive study areas. Nevertheless, further development is needed to improve the accuracy of the algorithms, as well as to derive geomorphological gully parameters (e.g., perimeter and volume) instead of pixel-level classification.

Find all details in the full paper:

The research of this paper is complementary to the project DEM_HYDR2024, funded by the Deutsches Zentrum für Luft- und Raumfahrt (DLR) for the used TanDEM-X datasets. Fieldwork campaigns needed for this research were funded by Integrated Land Management Institute (ILMI, grant number RY210400) and by the Department of Geo-Spatial Science and Technology at Namibia University of Science and Technology. Financial support was provided by the Deutsche Forschungsgemeinschaft for Open Access Publishing.
Methods of gully detection and monitoring are the core research subject of the PhD project of Miguel Orti on the development of gully identification and measurement methods combining remote sensing and crowdsourcing techniques.

Find out about the 3DGeo group’s work in further geomorphic settings, such as costal environments in the Netherlands, snow cover monitoring at Zugspitze mountain and rock glacier surface changes in the Austrian alps.

Vector-born diseases – such as Malaria, Dengue or Zika are serious health hazards in tropical regions. The outbreaks show high temporal and spatial variability. For example, the number of dengue cases in the state of São Paulo increased by 2,124% in the first 11 weeks of 2019 (up to March 16, 229,064 cases were reported), according to a survey by the Ministry of Health. It is important to monitor, model and predict such dynamics to be able to provide essential information for health administration units and NGOs.

Since most mosquitos have a short-distance flight range and a small dispersal rate, transmission of virus or parasites relies on the host’s movements. That means there should be strong connections between human’s movements and disease’s transmission. Therefore, it is necessary to capture human mobility pattern.

One way to extract human mobility pattern is to use geolocated Tweets. Compared with traditional commuting questionnaire-based surveys and phone call data, Twitter data is more convenient to work with. It has a relatively high spatial resolution and available both in near-real time as well as for past time periods. These mobility data will be later on combined with health report data to reveal the potential transmission patterns.

Tweets were collected from the Twitter free API, through which, we only can get 1% of total volume. Since Oct. 2018 to May 2019, in total 434,170 tweets from São Paulo city (Brazil) were collected, and visualized in figure 1a.

In order to align with our future processing, we choose the street block level as our minimal spatial unit. Street blocks were divided by the road network extracted from OSM (figure 1b).

Tweets were aggregated to street blocks and the number of movements of users estimated by sequential Tweets was calculated for all connected street blocks. This illustrates the strength of mobility connection between different street blocks (figure 1c and details in figure 1d). Large volumes of humans moving through an area intuitively increases the probability of contacting between an infected person and mosquitos present in the region. Using movements together with habitat quality from the perspective of mosquito, may help to estimate disease risk more precisely and to model potential outbreaks based on the strength of the connection between different street blocks (transmission chains).

Figure 1a: Visualized geo-located Tweets as yellow dots and sub-districts of São Paulo as background. Figure 1b: Enlarged graph from the center of São Paulo with satellite image as background and blue lines show street blocks derived from OSM, yellow dots are Tweets; Figure 1c: The strength of connection between street blocks based on sequential Tweet locations. Figure 1d: enlarged graph from the black rectangle in figure 1c, which shows the detailed connections between street blocks.

Today, June 5 is the #WorldEnvironmentDayWorld Environment Day is the United Nations day for encouraging worldwide awareness and action to protect our #environment. Above all, World Environment Day is the “people’s day” for doing something to take care of the Earth. That “something” can be local, national or global. This is a good opportunity to remind about the global OSM Climate Protection Map at Klimaschutzkarte.de by GIScience Heidelberg.

The OSM Global Climate Protection Map is based on user-contributed data from the free and open OpenStreetMap (OSM). It allows you both to share information about relevant geographic places as well as find out about topics related to climate protection and sustainability in your surrounding or area of interest. Some first relevant topics include map layers about for example renewable energy, different mobility forms as well as sustainable consumption and food etc. Click on one of the different categories to choose from and explore the related information layers on the map.

In the interactive online map, users worldwide can provide concrete references to facilities or offers that make it possible to make their own lifestyle climate-friendly - such as weekly markets with regional products, second-hand shops, car-sharing offers or facilities for the production of renewable energies in their area. Learn about how to contribute such (and more) information to OSM in the OSM Wiki

Better information is the first step for the right action. Also YOU can contribute to climate protection through a more sustainable lifestyle. The Climate Protection Map helps you to do this by showing where you can find e.g. healthy bio food or where to repair broken goods instead of throwing them to the waste and buying new stuff. Contribute your local knowledge about the five identified dimensions of climate protection to OpenStreetMap and raise the awareness of everyone on these important topics for society. Thanks to all OSM contributors for making this possible!

The five main dimensions provided by the map are at the moment the following. Each category can have several individual layers of information (see at the map):

- Energy Policy (e.g. Initatives and clubs)

- Energy Supply (e.g.Biogas-powerplants, gas storages, geothermal facilities etc.)

- Living and Construction (Proxy for house heating demand: Building Surface-Are to Volume Ratio, for Heidelberg only)

- Mobility and Traffic (e.g. Rent-a-bike-Stations, E-mobility charging stations, Car Sharing Places etc.)

Food and Consumption (e.g. Organic Supermarkets, Farm Shops, Repair Shops or Second-Hand Shops)

Individual regions can be compared to one another, and a greater awareness of citizens’ climate protection and energy needs is created.

The Klimaschutzkarte.de has been realized by GIScience Heidelberg in cooperation with ifeu Heidelberg (Institute for Energy and Environment Heidelberg) and was funded by Ministerium für Wissenschaft, Forschung und Kunst Baden-Württemberg, Germany.

From citizens for citizens: Share your local knowledge about the relevant topics in OSM and inform yourself about possibilities for more sustainable living!

Todays data production, maintenance, and use have changed in the last years.  While these tasks were reserved to professionals until a few years ago, the situation has changed.  This is no different in the geographical domain. Volunteers gather general information in Wikipedia and geographical information in OpenStreetMap.  Twitter users provide not only text snippets but in some cases also their current coordinates.  Whenever people interact in the production, maintenance, or use, they become part of a social process, leading to a new form of data sources.  Many terms used in this context (VGI, UGC, etc.) have some connotation by the way the are used.  Our journal article ‘Shared Data Sources in the Geographical Domain’ examines such social data sources in general without restricting to a certain domain.

We coin the term ‘Shared Data Sources’ as a generic umbrella term without any connotation that is often inherent to other terms: ‘A dataset or project is called a “Shared Data Source” (SDS) if its production, maintenance, and use are predominantly social processes.’ In addition, ‘we coin the term “Geographical Shared Data Sources” (GSDS) for referring to Shared Data Sources in the geographical domain.’ Feel free to do use these terms when referring to all these datasets.  Contributors and users share these datasets in various ways, they have become social!

Existing Shared Data Sources are often discussed in the context of Volunteered Geographic Information (VGI), Ambient Geographic Information (AGI), and Participatory Geographic Information (PGI).  We use these (proto)types to set Shared Data Sources in their mutual context.  The Triangle of Shared Data Sources (see above) is only one such example.  Our journal article contains some more such visualizations.

Mocnik, F.-B., Ludwig, C., Grinberger, A.Y., Jacobs, C., Klonner, C., Raifer, M. (2019): Shared Data Sources in the Geographical Domain—A Classification Schema and Corresponding Visualization Techniques ISPRS International Journal of Geo-Information 8(5), 242.

Related articles:

Mocnik, F.-B., Zipf, A., Raifer, M. (2017): The OpenStreetMap Folksonomy and Its Evolution. Geo-spatial Information Science 20(3), 219–230.

Mocnik, F.-B., Mobasheri, A., Griesbaum, L., Eckle, M., Jacobs, C., Klonner, C. (2018): A Grounding-Based Ontology of Data Quality Measures. Journal of Spatial Information Science 16, 1-25.

Mocnik, F.-B., Mobasheri, A., Zipf, A. (2018): Open Source Data Mining Infrastructure for Exploring and Analysing OpenStreetMap. Open Geospatial Data, Software and Standards 3(7).

Can you imagine how much sand is being moved on the beach in the course of a week? Did you ever observe truckloads of sand being transported on the beach in the absence of storms and bulldozers? It is hardly possible to estimate to the naked eye, but can be quantified with permanent terrestrial laser scanning (TLS).

This new paper investigates how the temporal interval of TLS acquisitions influences volume change observed on a sandy beach regarding the temporal detail of the change process and the total volume budget, on which accretion and erosion counteract. The study uses an hourly time series of TLS point clouds acquired over six weeks in Kijkduin, the Netherlands. Results of the hourly analysis are compared to those of a three‑ and six‑week observation period.

Results of change analysis for a three- and six-week period (left) and visualized as hourly time series of volume change for one location (right) (Anders et al. 2019)

Results of change analysis of sand on a beach for a three- and six-week period (left) and visualized as hourly time series of volume change for one location (right) (Anders et al. 2019)

For the larger, six-week period, a volume increase of 0.3 m³/m² is missed on a forming sand bar before it disappears, which corresponds to half its volume. Generally, a strong relationship is shown between observation interval and observed volume change. An increase from weekly to daily observations leads to a five times larger volume change quantified in total.

Another important finding is a temporally variable measurement uncertainty in the 3D time series, which follows the daily course of air temperature.

Will you be there? The research will be presented at the ISPRS Geospatial Week 2019 in the Change Detection session on Wednesday, 12th June 2019, 11:00 am - 12:30 am. The 3DGeo presents another research topic (Kumar et al. 2019) in the Machine & Deep Learning session.

Find all details in the full paper:

Anders, K., Lindenbergh, R. C., Vos, S. E., Mara, H., de Vries, S., and Höfle, B. (2019). HIGH-FREQUENCY 3D GEOMORPHIC OBSERVATION USING HOURLY TERRESTRIAL LASER SCANNING DATA OF A SANDY BEACH, ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci., IV-2/W5, 317-324, DOI: 10.5194/isprs-annals-IV-2-W5-317-2019

The paper is the result of fruitful cooperation within the research projects CoastScan (Dr. Sander Vos, Prof. Roderik Lindenbergh, and Prof. Sierd de Vries) and Auto3Dscapes under lead of Prof. Bernhard Höfle (3DGeo) and Dr. Hubert Mara (FCGL).

An important part of the research was conducted during the research visit of Katharina Anders with the group of Prof. Roderik Lindenbergh in the Geoscience & Remote Sensing department at TU Delft. We thank HGS MathComp for supporting the three-month research visit of PhD student Katharina Anders at TU Delft!

Methods of 4D geospatial data analysis are the core research subject of the Auto3Dscapes project on Autonomous 3D Earth Observation of Dynamic Landscapes. Find out about the 3DGeo group’s work in further geomorphic settings, such as snow cover monitoring at Zugspitze mountain and rock glacier deformation in the Austrian alps.

To stay updated, follow us on ResearchGate!

The PhD project Auto3Dscapes is funded by the Heidelberg Graduate School of Mathematical and Computational Methods for the Sciences (HGS MathComp), founded by DFG grant GSC 220 in the German Universities Excellence Initiative.

A paper investigating the relevance of (pre-calculated) features for 3D point cloud classification using deep learning was just published in the ISPRS Annals of Photogrammetry and Remote Sensing.

The study presents a non-end-to-end deep learning classifier for 3D point clouds using multiple sets of input features and compares it with an implementation of the state-of-the-art deep learning framework PointNet++. It is found that the classification accuracy improves by up to 33% when including normal vector features with multiple search radii and features related to spatial point distribution. The method achieves a mean Intersection over Union (mIoU) of 94%, outperforming PointNet++’s Multi Scale Grouping by up to 12%. The paper presents the importance of multiple search radii for different point cloud features for classification in an urban 3D point cloud scene acquired by terrestrial laser scanning.

The study uses point clouds from the semantic3D dataset, a labelled 3D point cloud data set of geographic scenes. The figure below shows the labelled 3D point cloud from the paper as ground truth and a result of the deep learning classifier for one of the feature sets (cf. Kumar et al. 2019).

Non-end-to-end deep learning point cloud classification (Kumar et al. 2019)

Result of non-end-to-end deep learning point cloud classification (Kumar et al. 2019)

Wonder how misclassifications in the result can be explained? Find all the details in the paper:

Kumar, A., Anders, K., Winiwarter, L., and Höfle, B. (2019). FEATURE RELEVANCE ANALYSIS FOR 3D POINT CLOUD CLASSIFICATION USING DEEP LEARNING, ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci., IV-2/W5, 373-380, DOI: 10.5194/isprs-annals-IV-2-W5-373-2019

The research will be presented at the ISPRS Geospatial Week 2019 at University of Twente in Enschede (NL) in the Machine & Deep Learning session on Wednesday, 12th June 2019, 9:00 am - 10:30 am.

Will you be there?

The 3DGeo presents another research topic (Anders et al. 2019) from the Auto3Dscapes project in the Change Detection session.

Older Posts »