Feed on
Posts
Comments

Recently a consultancy and development agreement about OpenStreetMap Analytics Development has been reached with the World Bank in the context of the Open Cities Africa project and the Global Facility for Disaster Reduction and Recovery (GFDRR) Open Data for Resilience Initiative (OpenDRI).

The main objective of this consultancy is to develop and implement new functionalities for OpenStreetMap Analytics (OSMA). OSMA is an open source platform that allows interactive visualizations and analysis of OSM map data.

Through the Open Cities Africa project, GFDRR will be supporting community mapping project in urban areas across Africa and OSMA will be leveraged to document mapping progress and provide insights into the data and its contributors. All functionalities developed will be open source and made available to the broader OpenStreetMap community.

The specific objectives include participating in the design phase of new OSMA functions, researching implementation options and feasibility, provide input for a design firm to produce final mockups. Most importantly it includes the  implementation of new functions as well as contributing and documenting code publicly on Github.

Working closely with GFDRR staff and OSM community partners, specific features are being scoped out to expand the number of analytical tools and datasets available in OSMA. A survey of mapping partners’ needs has been conducted and provides some reference to the range of analytical functions that will be developed.

Also the current version of osm-analytics.org is already hosted on servers at the Heidelberg Institute for Geoinformation Techology (HeiGIT).

The Global Facility for Disaster Reduction and Recovery (GFDRR) is a partnership of the World Bank, United Nations, major donors and recipient countries under the International Strategy for Disaster Reduction (ISDR) system to support the implementation of the Hyogo Framework for Action (HFA).

In 2011, Global Facility for Disaster Reduction and Recovery (GFDRR) launched the Open Data for Resilience Initiative (OpenDRI) to apply the concepts of the global open data movement to the challenges of reducing vulnerability to natural hazards and the impacts of climate change.

Building on the success of the global OpenDRI, its work on Open Cities projects in South Asia, and GFDRR’s Code for Resilience, in 2018 Open Cities Africa will be carried out in selected cities in Sub-Saharan Africa to engage local government, civil society, and the private sector to develop the information infrastructures necessary to meet 21st century urban resilience challenges.

The project is already going on with first prototype results. Stay tuned for further information.

Satellite images are widely applied in humanitarian mapping which labels buildings, roads and so on for humanitarian aid and economic development. However, the labeling now is mostly done by volunteers. In a recently accepted study, we utilize deep learning to solve humanitarian mapping tasks of a mobile software named MapSwipe. The current deep learning techniques e.g., Convolutional Neural Network (CNN) can recognize ground objects from satellite images, but rely on numerous labels for training for each specific task. We solve this problem by fusing multiple freely accessible crowdsourced geographic data, and propose an active learning-based CNN training framework named MC-CNN to deal with the quality issues of the labels extracted from these data, including incompleteness (e.g., some kinds of object are not labeled) and heterogeneity (e.g., different spatial granularities). The method is evaluated with building mapping in South Malawi and road mapping in Guinea with level-18 satellite images provided by Bing Map and volunteered geographic information (VGI) from OpenStreetMap, MapSwipe and OsmAnd.

In comparison with previous VGI and deep learning work, the advantage of our study lies in (i) combining multiple VGI data with technical solutions to deal with the quality issues (i.e., incompleteness and heterogeneity), and (ii) empirically studying the deep learning method’s application in the humanitarian mapping software MapSwipe with a machine-volunteer collaboration mechanism.

The results based on multiple metrics including Precision, Recall, F1 Score and AUC show that MC-CNN can fuse the crowdsourced labels for higher prediction performance, and be successfully applied in MapSwipe for humanitarian mapping with 85% labor saved and an overall accuracy of 0.86 achieved. With the evaluation on real world data in Africa, we found that

(i) combining multiple VGI data significantly outperforms one single VGI data because of the increased sample diversity,

(ii) training of MC-CNN needs sample sets with large enough size,

(iii) MC-CNN can achieve robust learning for different different CNN architectures including LetNet, AlexNet and VggNet,

(iv) MC-CNN saves a large part of labor but keeps high overall accuracy when it is applied in MapSwipe with the machine-volunteer collaboration mechanism.

These findings will benefit the deep learning-based exploiting of VGI data for humanitarian mapping. This work has been conducted in the context of the ongoing DeepVGI project at HeiGIT. Stay tuned for future updates and new results!

Chen, J., Y. Zhou, A. Zipf and H. Fan (2018, accepted):  Deep Learning from Multiple Crowds: A Case Study of Humanitarian Mapping. IEEE Transactions on Geoscience and Remote Sensing (TGRS). accepted. vol/pp pending.


Related work:

Chen, J., Zipf, A. (2017): Deep Learning with Satellite Images and Volunteered Geographic Information (VGI). In: Karimi, H. A. and Karimi, B. (eds.): Geospatial Data Science Techniques and Applications. Chapter 3. pp. 63-78. crc press. Taylor & Francis.

The concept of place recently gains momentum in GIScience. In some fields like human geography, spatial cognition or information theory, this topic already has a longer scholarly tradition. This is however not yet completely the case with statistical spatial analysis and cartography. Despite that, taking full advantage of the plethora of user-generated information that we have available these days requires mature place-based statistical and visualization concepts. A recently accepted and published paper for GIScience 2018 conference in Melbourne contributes to these developments: We integrate existing place definitions into an understanding of places as a system of interlinked, constituent characteristics. Based on this, challenges and first promising conceptual ideas are discussed from statistical and visualization viewpoints. One major challenge is to find suitable units upon which statistical analyses of places can be conducted. Conceptual spaces have been identified as one promising way to define such units, though an in-depth harmonization of this framework with places still needs to be done in future work. Further, platial counterparts to important spatial-statistical concepts must be formulated in order to develop a valid and rigorous statistical theory of places. It is not yet clear to what extent data taken from user-generated feeds is truely platial. Since data is a crucial ingredient to achieving insights on places, this is one of the major empirical steps to be undertaken in the near future. In terms of visualizing places, the major issues with current approaches include wrong spatial impressions created through interpolation techniques, the problem of displaying multifaceted place-based information at once, and the combination of different subjective places in one map. However, the proposed example using micro diagrams has shown first promising results for the presentation of multidimensional, qualitative information together with the spatial outline of places in a conceivable way.

Westerholt, R., Gröbe, M., Zipf, A. and Burghardt, D. (2018): Towards the statistical analysis and visualization of places. 10th International Conference on Geographic Information Science, Melbourne, Australia, DOI: 10.4230/LIPIcs.GIScience.2018.63.

Related work on PLATIAL analytics will be presented and discussed at our PLATIAL 2018 Workshop in Heidelberg. 20.-21.Sept. 2018. Join the Discussion!

Over the past few years, the Missing Maps approach has repeatedly proved its potential for humanitarian assistance and disaster management. While the project was launched by only four organizations, there are now 17 member organizations in Missing Maps, and nearly 60,000 mappers. The Red Cross and Red Crescent Movement have recognized the potential of the Missing Maps project and its approach, acknowledging the crucial contributions of geographic information in driving disaster responses and recovery as well as preparedness. Together with our colleagues from the British Red Cross and German Red Cross, we published a perspective article, which gives an overview on the Missing Maps approach and its potential within the Red Cross and Red Crescent Movement:

Scholz, S.; Knight, P.; Eckle, M.; Marx, S.; Zipf, A. (2018): Volunteered Geographic Information for Disaster Risk Reduction—The Missing Maps Approach and Its Potential within the Red Cross and Red Crescent Movement. Remote Sens. 2018, 10(8), 1239, doi: 10.3390/rs10081239.

On our last day in the Ötztal Alps, we had an exciting excursion from Obergurgl to Ramolhaus on 3006 m a.s.l. On our way up we could directly explore glacial history by passing the historical extents and related moraines of the retreating Gurgler glacier. The students also learned about geoarcheology and settlement history in the Ötztal valley, where people settle for more than 9000 years.

Moreover, we had the full view on the active rock glacier Äußeres Hochebenkar, which has been investigated for several years by the 3DGeo group.

The excursion rounded up a very interesting field trip where the students learned a lot about 3D surface and geophysical subsurface geodata acquisition in glacial and periglacial process areas.

The students would like to thank Prof. Bernhard Höfle, Dr. Stefan Hecht, and Katharina Anders for providing a wide range of methodical and content-related input and for organizing this great practical field trip in an impressive high-mountain environment.

We are blogging live from Ramolhaus at 3006 m a.s.l.

After a 4-hours hiking tour with 1100 height meters we are now enjoying the impressive view over glaciers with “Kaiserschmarren” for lunch.

Over the past few months the openrouteservice team has worked on a new developers dashboard in order to make the registration and usage of the API more easy and accessible.

Now the developers using the openrouteservice API will experience a more intuitive and responsive interface and a with more polished better look and feel to view their quota and tokens. To achieve this modular solution we have created a single page application using the novel framework Vue.js developed by the team around Gitlab.com .

What’s new?

  • User Authentication
  • Tokens listing, removal, creation
  • Tokens usage chart
  • Profile view/update
  • One-click Login and or Sign-up with Github Credentials
  • More intuitive and smart forms validation
  • Invisible captcha
Developer Dashboard login:

On the last day of fieldwork in the rotmoos valley two groups (terrestrial laser scanning (TLS) and photogrammetry) set off to the rotmoos glacier again. A second TLS dataset was aqcuired
which enables the students to perform a change detection or deformation analysis.

The electrical resistivity tomography (ERT) group futher explored the pre-Quaternary relief in the valley and measured another profile in the area of a ground moraine to verify previous measurements and to differentiate sediment types and process areas. The various marmots living in and around the study site are not only cute but the excavation material of their burrows also provide insights into the spatially varying composition of sediments along the ERT measurement profiles (see pictures below).

After fieldwork the students climbed the Schönwieskopf mountain to enjoy the beautiful view to Obergurgl and Rotmoos valley.

Every evening the acquired data are processed in the “sky-lab” in the Obergurgl University Center. The photogrammetry group, for example, performs a 3D reconstruction of glacial landforms which have been captured in the field with smartphone cameras.

Moreover, the 3DGeo team extended the multi-temporal terrestrial LiDAR dataset of the rock glacier Äußeres Hochebenkar in the last days. Additionally, the first UAV LiDAR point cloud of this rock glacier was acquired by Martin Rutzinger and Magnus Bremer from the Institute for Interdisciplinary Mountain Research (IGF) and Innsbruck University. They flew with a RiCOPTER UAV drone with a RIEGL VUX-1LR laserscanner. These datasets provide a valuable basis for further research on the rock glacier.

We are excited to announce that our new HeiGIT website is live! Visit us at www.heigit.org.

Today was the second day of field work in the Rotmoos valley in the Ötztal Alps for our 16 students. The hot sun challenged the data acquisition but the impressive landscape makes up for every effort.

The terrestrial laser scanning group climbed up the valley flanks to acquire a high resolution point cloud from an end moraine of the Rotmoos glacier from the year 1850.
The electrical resistivity tomography (ERT) group measured a profile along the transition from the valley flank to the valley floor next to the end moraine. The measurement profile was aimed to track the bedrock and to determine the thickness of valley backfilling sediments.
The photogrammetry group took multiple photos from different landforms (e.g. rôches moutonnées and debris cones) in order to perform a 3D reconstruction of these objects. Moreover, the students learned how to set up a real time kinematic global navigation satellite system (RTK GNSS) which is used to survey global coordinates (e.g. of the ERT profile).

For more information about this practical field course read our blog post from yesterday.


We will keep you updated with daily posts - stay tuned.

Older Posts »