Thursday 29 Aug 2019
12:30 pm - 2:00 pm Posters Session
A Machine Learning Pipeline Articulating Satellite Imagery and OpenStreetMap for Road Detection
Mayra Zurbaran (Politecnico di MIlano)National Theater
Satellite imagery from earth observation missions enable processing of big data for gathering information about the world. Automatizing the creation of maps that reflect ground truth is a desirable outcome that would aid decision makers to take adequate actions in alignment with the United Nations Sustainable Development Goals. In order to harness the power that the availability of the new generation of satellites enable, it is necessary to implement techniques capable of handling annotations for the massive volume and variability of high spatial resolution imagery for further processing. Demir et. al, 2018 states that satellite images are only recently gaining attention from the computer vision community for map composition ; however, the availability of public datasets for training machine learning models ML for image segmentation plays an important role for scalability. This work focuses on bridging remote sensing and computer vision by providing an open source based pipeline for generating machine learning training datasets for road detection in an area of interest. The proposed pipeline addresses road detection as a binary classification problem using road annotations existing in OpenStreetMap OSM for creating masks. The OSM vector data is preprocessed using a buffer for covering the road with the tagged width value to generate annotated images and the corresponding overlapping satellite image ready to be used on a machine learning model as arrays. For testing the output, a TensorFlow neural network available in https github.com mahmoudmohsen213 airs was used. For this case study, Planet and Sentinel-2 images are used for creating training datasets for road detection in Kenya. Existing ML models lack accuracy in predictions due to the challenges posed by the similarity between the landscape and urban areas in this region, seasonality of roads, and limited availability of high-resolution imagery. The pipeline takes into consideration OSM features present in this region to produce masks that match as closely as possible the ground truth of the area of interest, considering an average road width measured from the images for mapped roads without this metadata. The pipeline is written in python with pyQGIS and will be made available as open source. Demir, I., Koperski, K., Lindenbaum, D., Pang, G., Huang, J., Basu, S., Raska, R. 2018 . DeepGlobe 2018 A challenge to parse the earth through satellite images. In IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops Vol. 2018 June, pp. 172 181 . http doi.org 10.1109 CVPRW.2018.00031
An open-source mobile geospatial platform for promoting climate-smart livelihood-landscape systems in Fiji and Tonga.
Kevin P Davies (School of Geosciences, Faculty of Science, University of Sydney, NSW 2006, Australia) , John Duncan (UWA School of Agriculture and Environment, The University of Western Australia, WA 6009, Australia) , Nathan Wales (School of Geography, Earth Science and Environment, The University of the South Pacific, Suva, Fiji) , Renata Varea (School of Geography, Earth Science and Environment, The University of the South Pacific, Suva, Fiji) , Renata Varea (School of Geography, Earth Science and Environment, The University of the South Pacific, Suva, Fiji) , Helena Shojaei (UWA School of Agriculture and Environment, The University of Western Australia, WA 6009, Australia) , Eleanor Bruce (School of Geosciences, Faculty of Science, University of Sydney, NSW 2006, Australia) , Bryan Boruff (UWA School of Agriculture and Environment, The University of Western Australia, WA 6009, Australia) , Eloise Biggs (UWA School of Agriculture and Environment, The University of Western Australia, WA 6009, Australia)National Theater
Communities in Fiji and Tonga rely on landscape services to support a variety of livelihoods. These communities are increasingly vulnerable to climate e.g. increasing cyclone occurrence and intensity and environmental e.g. mining and deforestation stressors. Within these landscape systems, accurate and timely monitoring of human-climate-environment interactions is important to inform landscape management, land use policies, and climate-smart sustainable development. Data collection and monitoring approaches exist to capture landscape-livelihood information such as surveys, participatory GIS PGIS , and remote sensing. However, these monitoring approaches are challenged by data collection and management burdens, timely integration of databases and data streams, aligning system requirements with local needs, and socio-technical issues associated with low-resource development contexts. Such monitoring approaches only provide static representation of livelihood-landscape interactions failing to capture the dynamic nature of vulnerabilities, and benefit only a small user base. We present a prototype of a mobile, open-source geospatial tool being collaboratively developed with the Ministries of Agriculture in Fiji and Tonga and local stakeholders, to address the above shortcomings of PGIS and other environmental monitoring and data sharing approaches. The tool is being developed using open-source mobile GIS technologies following a formal ICT for Development ICT4D methodology. We discuss the results and challenges of all the phases of the ICT4D methodology which involves multiple landscape stakeholders across the two Small Island Developing States. Based on outcomes from the ICT4D user requirements analysis, we produced a prototype open-source mobile geospatial data collection, analysis and sharing tool. New dynamic spatial data layers related to landscape use and climate were specifically developed for use in the tool. We present the functionality of the tool alongside the results of field-testing with communities in the Ba Catchment, Fiji and Tongatapu, Tonga.
Analysing Geographical Access to Health Care in Nigeria using QGIS
This study explores geographical access to healthcare delivery and facilities of sampled households in Nigeria based on spatial access using Availability number of local service points available and Accessibility -travel impedance distance, types of transport, mode of transport, costs of transport, time taken to determine to healthcare. It presents a workflow of using open and free software from data collection to building the database of health facilities and household characteristics using Post GIS and cost distance analysis using SAGA GIS plug in of QGIS. Bringing in geographical variables such as DEM, road condition and types and socio economic variables, findings showed that while availability of healthcare facilities according to age is a challenge, households have not always used the nearest hospital that is available and physically accessible to them, but rather travel further. The costs of these travels are revealed. Further investigation into barriers into showed other socio - cultural issues such as costs, types and availability of health insurance, use of employer s health facilities as larger issues considered beyond of spatial access. The patterns of these health care usage and access in the sampled household are presented.
Analysis of EGNOSS ionospheric model’s impact on the integrity level in the Central and Eastern Europe region
Balázs Lupsic (Budapest University of Technology and Economics Faculty of Civil Engineering Department of Geodesy and Surveying) , Bence Takács (Budapest University of Technology and Economics Faculty of Civil Engineering Department of Geodesy and Surveying)National Theater
The demand of Global Navigation Satellite System in safety related applications has rapidly increased the last few years. The foreseeable release of self-driving cars are already showing the importance of the integrity concept of satellite based navigations. Correction services as EGNOS can actively improve the integrity assurance build up. The paper s aim is the examination of the EGNOS ionospheric model in regard of integrity. The focus is on the Central Eastern European area where the users can get close to the east edge of the EGNOS coverage area. We processed the available recorded broadcasted EGNOS data from 2018 and compared the ionospheric corrections performance with a profoundly credible model. The paper presents the basic statistical properties of the comparison with the highlights of deviances, which could endanger the navigation integrity. The paper has an additional focus on how the quality of the EGNOS ionosphere model can influence the protection level in the eastern region. Satellites with low elevation angle may be out of the EGNOSS coverage area and the absence of these transmitters can negatively influence the protection level. The paper shows the quality and quantity of the above mentioned negative impact with the help of real life and simulated data. At the end there is a proposal for simple strategies which could mitigate the degradation of the protection level at the eastern edge of the EGNOS ionosphere coverage area due to lack of ionospheric correction.
Citizen Science for water quality monitoring applying FOSS
Stefan Jovanovic (Politecnico di Milano) , Maria Antonia Brovelli (Politecnico di Milano) , Daniela Carrion (Politecnico di Milano)National Theater
In the framework of SIMILE Sistema Informativo per il Monitoraggio Integrato dei Laghi insubrici e dei loro Ecosistemi, Information System for the Integrated Monitoring of insubric lakes and of their ecosystems Interreg Italy-Switzerland project, the citizens contribution to monitor the quality of lakes water has been envisaged. In the initial phase of this research, state of art of Citizen Science and water quality monitoring was investigated. The analysis of past and current projects, governed by different organizations and communities, pointed out a variety of tasks that can be accomplished by citizens. In one of the cases, participants were recruited to collect images where algae blooms are present potentially indicating eutrophication , while in some other they were involved in determining waters chemical composition. In these studies, as well as in others that were analyzed, authors stressed suitability of smartphones for the fulfilment of various assignments given to citizens image capturing, geotagging of collected data, filling and submitting form . Further on, in this research different smartphone applications for water quality monitoring were tested through Citizen Science project. The differences between applications, in terms of additional required tools devices, observables and obtained outputs, were analyzed and reported. Besides that, the accessibility of stored published outputs as well as their availability whether they are free and open or not were evaluated. Applications tested in this research are generating particular outputs, targeting different interest groups like stakeholders, policymakers, researchers, etc. dealing with water quality monitoring. Despite the fact that many applications for water quality monitoring are freely available, none of them is open source. Hence, this paper is proposing the design of a new application that will be free and open source, addressing not just users but also developers giving them a possibility for customization and improvement.
Cross-border open data sharing: GIOCOnDA project
Juan Fernando Toro (Politecnico di Milano - DICA - Geodesy and Geomatics Section) , Daniela Carrion (Politecnico di Milano - DICA - Geodesy and Geomatics Section) , Alberta Albertella (Politecnico di Milano - DICA - Geodesy and Geomatics Section) , Maria Antonia Brovelli (Politecnico di Milano - DICA - Geodesy and Geomatics Section)National Theater
The possibility to exploit open data as a generator of value has brought the need for international initiatives to ease the data interoperability. Data sharing can be relevant for the economic growth of different markets within a territory, especially when it requires the communication on a cross-border territory so that asymmetries of information are reduced and policymaking becomes more efficient. The GIOCOnDA Gestione Integrata e Olistica del Ciclo di vita degli Open Data, Integrated and holistic management of Open Data life cycle Interreg Italy-Switzerland project surges as a proposal for compiling the open data sources within the Insubria Region under a single infrastructure and, with this, to stimulate the territory economy. The territory of interest comprehends the cross-border region between Italy and Switzerland, including Lombardy Region and Canton of Ticino. This territory relies on tourism as one of the main economic sources, for this reason, the study is focused on it. A central issue for this purpose is the incompatibility and underuse of the current local governments open data catalogs. In addition, the potentially very rich content of non-authoritative, publicly contributed datasets, such as OpenStreetMap, must be considered and, after a quality check, integrated into the system. The delivery of a common framework for data sharing, where interoperable data are available for the Insubria Region, is required. At the initial stage of the GIOCOnDA project the datasets matching the needs of the tourism sector are collected. Firstly, by understanding the requirements for information being considered as useful by the end-users of the tourism sector Public Administrations, stakeholders, citizens . Secondly, by revising the data catalogs type of data, availability, typology of dissemination and possible extensions of the current information with geospatial features, particularly in terms of mobility as for example, transportation means, schedules, etc. . Inconsistencies were singled out in the open data classification between Italy and Switzerland, with respect to the case studies of the project, Tourism and Mobility. A proposal of data categorization by typology, according to a common model, consistent with the existing standards, has been made. Authoritative and non-authoritative data have been compared to allow an effective integration. The analyses on the datasets have been performed with open source software, such as QGIS. The objective to be pursued is to open useful data, to guarantee the monitoring of data quality, allowing more effective usability and reuse and sustainability over time. In addition, it can become a reference study case for the first steps on data compilation for interoperability on an international context.
Effect of the Location of health Facilities on Skilled Deliveries in Achieving Universal Health Coverage in Kisumu County, Kenya
Caroline Akoth (Women in GIS, Kenya) , Kevin Rombosia (County Government of Kisumu)National Theater
Introduction Kisumu County is one of the four counties in Kenya that have been selected to pilot universal health care UHC . Key UHC index measures include the proportion of skilled deliveries conducted and health facilities access World Health Organization, 2015 . Households should be located within 5 kilometers access to medical services Kenya, 2017 . In Kisumu County, the extent of primary healthcare coverage and how this accessibility affects skilled delivery is also not known. We thus sought to describe health facility accessibility and its effect on skilled delivery in Kisumu County in line with the pillars of UHC. Methods This was a retrospective chart review. The facilities are categorized by type and level of care. Private healthcare facilities and those that do not conduct deliveries were excluded in this study. Secondary data proportion of skilled deliveries for 156 health facilities conducted in 2016 was mined from the District Health Information System. Healthcare physical services accessibility was represented using a 5km radius fixed distance buffering around health facilities and the health facilities distances to the nearest major roads. Voronoi polygons were also generated to represent coverage by facilities. Bivariate analysis of distance to nearest roads and skilled deliveries done using simple linear regression was then done. Results The mean skilled delivery was 45.8 Median 45.8 , Range 0 to 358 and IQR 48.6 . There exist 4 pockets of underserved areas in Nyando, Nyakach and Muhoroni sub counties measuring 21 kms2, 52 kms2, 60 kms2, 65 kms2 and 94 kms2 respectively. Regarding the effects of a health facilities distance from the nearest road to skilled deliveries conducted, the R2 value was 0.02 with a negative slope following a linear regression analysis of the two variables. Discussion The underserved areas are mostly located away from major roads. The mean skilled delivery was lower than the national target of 80 . However some facilities exceeded 100 . This can be explained by in referrals that cause such facilities to exceed their projected workloads. The distance of a health facility to the nearest major road is inversely proportional to the skilled deliveries conducted. However, this comparison is weak in establishing such an effect. There are pockets of underserved areas in Kisumu County regarding physical access to health care services. There is a need to conduct a further study on how reproductive, maternal, neonatal and child health indices, communicable diseases and non-communicable diseases indices, service capacity and health care access indices affect the overall UHC index in Kisumu County.
Evaluating student motivation and productivity during mapathons
Serena Coetzee (Centre for Geoinformation Science, Department of Geography, Geoinformatics and Meteorology, University of Pretoria) , Victoria Rautenbach (Centre for Geoinformation Science, Department of Geography, Geoinformatics and Meteorology, University of Pretoria) , Cameron Green (Centre for Geoinformation Science, Department of Geography, Geoinformatics and Meteorology, University of Pretoria)National Theater
When disaster strikes developing countries, a lack of geographic data is often a hindrance to first response and relief operations. To address this lack of geographic data, remote mapping and especially mapathons have played an important role in collecting geographic data in OpenStreetMap OSM that can be used to plan activities in areas effected by disaster or other humanitarian efforts. During mapathons, volunteers from various backgrounds get together to map a specific area using satellite imagery or aerial photographs. The expertise and motivation of these volunteers generally differ. Even though the geographic data in OSM is invaluable in cases where little to no data is available, the quality of the data collected during mapathons is often questioned. In this paper, we present our results from an evaluation of university students motivation for participating in mapathons and their productivity i.e. how much data they contributed . To achieve our aim, we hosted four mapathons for final year university students where the participants were asked to complete a short questionnaire to determine their motivations and personal opinions of the mapathon. Afterwards, the productivity for two mapathons were evaluated. Final year students enrolled in a geoinformatics module were offered extra credit for participating in the mapathons. As a result, the majority of the students participated in all four mapathons and the answers did not differ significantly between the four mapathons. One of the main reasons mentioned, apart from extra credit, was that the participants felt a sense of humanitarianism through contributing to communities in need by assisting. Additionally, the social aspect also came through with a large percentage of the participants indicating that mapathons are fun and that they learned something new, for example by improving their digitizing skills or that humanitarian organizations need help. Participants also indicated that the tools i.e. OSM and iD editor were easy to use, but that the imagery is sometimes not good enough due to cloud coverage. The general productivity for two mapathons was evaluated and we found that with more experience the participants were generally more productive. A further step was taken by investigating five individuals productivity. It was clear that their productivity increased, and that they made fewer errors during the subsequent mapathons. The results from this evaluation provided insight and knowledge that could assist mapathon organisers to create a more productive environment for participants with the hopes of encouraging them to produce high quality data. The feedback from students was clear that if they receive information about the aim of a mapathon and why the data is important, they are more motivated to produce high volumes of quality data.
GeoCMS : Towards a Geo-tagged Media Management System
Byungcheol Kang (Turbosoft) , Wei Ding (Kunsan National University) , Donggeon Lee (Kunsan National University) , Jemin Song (Turbosoft) , Kwang Woo Nam (Kunsan National University)National Theater
A large amount of daily geo-tagged media data generated by user s smart phone, mobile device, dash cam and camera. Geo-tagged media, geovideos and geophotos, can be captured with spatial temporal information such as time, location, visible area, camera direction, moving direction and visible distance information. Due to the increase in geo-tagged multimedia data, the researches for efficient managing and mining geo-tagged multimedia are newly expected to be a new area in database and data mining. This paper proposes a geo-tagged media management system, so called GeoCMS Geo Contents Management System . GeoCMS is a new framework to manage geo-tagged media data on the web. Our framework supports various types which are for moving point, moving photo - a sequence of photos by a drone, moving double and moving video. Also, GeoCMS has the label viewer and editor system for photos and videos. GeoCMS have been developed as an open source system. You can find the repository in https github.com turbosoft GeoCMS.
Geometry viewer for pgAdmin4: a Process guided by the Google Summer of Code
Xuri Gong (Institute of Remote Sensing and Geographic Information Systems, School of Earth and Space Sciences, Peking University, China) , Frikan (Verge Technologies, South Africa) , Victoria Rautenbach (Centre for Geoinformation Science, Department of Geography, Geoinformatics and Meteorology, University of Pretoria, South Africa)National Theater
NATURAL HERITAGE MANAGEMENT AND PROMOTION THROUGH FREE AND OPEN SOURCE SOFTWARE: A PRELIMINARY SYSTEM DESIGN FOR THE INSUBRIPARKS PROJECT
Maria Antonia Brovelli (Politecnico di Milano) , Massimiliano Cannata (IST-SUPSI) , Vittorio Terza (Parco Regionale Spina Verde) , Daniele Oxoli (Politecnico di Milano)National Theater
The management and promotion of peculiar territories such as protected areas and natural parks have emerged as critical tasks to unpin their tourism potential while ensuring sustainable exploitation of their landscape and environmental resources. These tasks are often hindered by fragmented political contexts to which a same protected area may be subjected. This may prevent optimal management due to competitive actions as well as weak cooperation among the decision makers. With this in mind, we introduce here the INSUBRIPARKS project, an Interreg funded project aiming at the harmonisation of management and promotion practices of natural parks along the cross-border area between Italy and Switzerland. The Interreg program is an initiative of the European Union aiming at promoting investments, innovation and implementation effort to boost cooperation among regional and local governments across the continent to ensure sustainable impacts for people and territories. According to this, the INSUBRIPARKS project objective is to develop a network of tourism experiences and facilities within the cooperation among project partners. The project encompasses multiple actions including the provision of physical infrastructure and the development of a standardized IT infrastructure. The latter is designed to improve information generating and consuming among project partners and stakeholders. The target content ranges from geospatial datasets - including tourism landmarks and facilities - to information from the crowd such as social media posts and reviews. Environmental sensor observations, namely visitor counts, will be also considered. These will be collected by a low-cost sensors network dispatched around touristic hotspots within the project areas. The IT infrastructure will include three main components. A server-side will be implemented to dynamically store, organize, and expose both geospatial data and non-spatial content. The data endpoint will be implemented considering cutting edge FOSS solutions such as GeoServer and istSOS. Upon the latter component, a custom dashboard enabling data management and processing will be created to provide a Business-to-Business platform to the project partners. Finally, a Web client will provide users with access and interaction with the information layers. The client will take advantage of standard FOSS Web mapping libraries such as OpenLayers and it will be enriched with custom modules allowing a dynamic interaction with the data endpoint. Alongside the preliminary design of the system architecture, both use cases and user requirements are discussed. Issues connected to critical data collection and harmonization are outlined. The assets provided by the use of FOSS for the development of the standardized IT infrastructure of the INSUBRIPARKS project is discussed together with the underlying benefits deriving from the co-creation of best management practices by means of open and shared software platform.
Object-based Image Analysis for historic maps classification
Stefano Gobbi (University of Trento, Fondazione Edmund Mach, Mountfor RC) , Paolo Zatelli (University of Trento) , Clara Tattoni (University of Trento) , Marco Ciolli (University of Trento, C3A) , Nicola La Porta (Fondazione Edmund Mach, Mountfor RC)National Theater
Stefano Gobbi 1,2,3 , Paolo Zatelli 1 , Clara Tattoni 1 , Marco Ciolli 1 , Nicola La Porta 2,3 . 1 Universit degli Studi di Trento, Dipartimento di Ingegneria Civile, Ambientale e Meccanica, Trento, Italy 2 Fondazione Edmund Mach, San Michele all Adige TN , Italy 3Mountfor Research center, San Michele all Adige TN , Italy Heritage maps represent fundamental information for the study of the evolution of a region, especially in terms of landscape and ecologic features. Historic maps present two kinds of hurdle before they can be used in a modern GIS they must be geometrically corrected to correspond to the datum in use and they must be classified to exploit the information they contain. This study deals the latter problem the Historic Cadaster Map, created between 1851 and 1861, for the Trentino region in the North of Italy is available as a collection of maps in the ETRF89 UTM 32N datum. The map is a high resolution scan 230 DPI, 24 bit of the original map and has been used in several ecological studies, since it provides detailed information not only about land property but also about land use. In the past the cadaster map has been manually digitized and for each area a set of attributes has been recorded. Since this approach is time consuming and prone to errors, automatic and semi-automatic procedures have been tested. Traditional image classification techniques, such as maximum likelihood classification, supervised or un-supervised, pixelwise and contextual, do not provide satisfactory results for many reasons map colors are very variable within the same area, symbols and characters are used to identify cadaster parcels and locations, lines, drawn by hand on the original map, have variable thickness and colors. The availability of FOSS tools for the Object-based Image Analysis OBIA has made possible the application of this technique to the cadaster map. This paper describes the use of GRASS GIS and R for the implementation of the OBIA approach for the supervised classification of the historic cadaster map. It describes the determination of the optimal segments, the choice of their attributes and relevant statistics, and their classification. The result has been evaluated with respect to a manually digitized map using Cohen s Kappa and the analysis of the confusion matrix. The result of the OBIA classification has also been compared to the classification of the same map using maximum likelihood classification, un-supervised and supervised, both pixelwise and contextual. The OBIA approach has provided very satisfactory results with the ability to automatically remove the background and symbols and characters, creating a reuse classified map. This study highlights the effectiveness of the OBIA processing chain available in the FOSS4G ecosystem, and in particular the added value of the interoperability between GRASS GIS and R.