Veranstaltungsprogramm

Eine Übersicht aller Sessions/Sitzungen dieser Tagung.
Bitte wählen Sie einen Ort oder ein Datum aus, um nur die betreffenden Sitzungen anzuzeigen. Wählen Sie eine Sitzung aus, um zur Detailanzeige zu gelangen.

 
Sitzungsübersicht
Sitzung
B21: GI in Urban Contexts
Zeit:
Freitag, 05.07.2019:
9:00 - 10:30

Chair der Sitzung: Angela Hof
Ort: B - Blauer Hörsaal
HS402

Zeige Hilfe zu 'Vergrößern oder verkleinern Sie den Text der Zusammenfassung' an
Präsentationen

Parametric (Geo)Design for Test Planning

Christoph Schaller

Berner Fachhochschule, Schweiz

The revised Swiss Spatial Planning Act pursues inward settlement development to slow down urban sprawl and protect arable land. This requires quality-oriented and sustainable internal densification. The Geodesign Framework by Steinitz can support planning processes with public participation, where models and visualizations help to convey complex interrelationships to stakeholders. This paper presents a Geodesign based process model integrating GIS and Parametric Design, that allows to quantify and communicate effects on internal densification caused by changes to building regulations. An overview of the model and the most important results from its verification are presented. They indicate that the model approximates real interrelations well and is a suitable basis for further work.



A 3D Spatial Data Framework for Urban Land and Property Management

Abbas Rajabifard, Jihye Shin, Behnam Atazadeh, Mohsen Kalantari.

The Centre for Spatial Data Infrastructures and Land Administration, Department of Infrastructure Engineering, The University of Melbourne

Over the last years, the unprecedented urbanization has fostered the rapid development of multi-story buildings and infrastructure facilities, resulting in spatial and functional complexities in cities. Land and property information plays an important role in a wide range of applications in land administration and management in rapidly growing cities. However, the current fragmented practice relied on 2D-based representation does not provide a reliable and accurate legal description of underground and aboveground properties as a foundation of making evidence-based decisions in support of economic prosperity, human activities and the public safety in urban areas. In this paper, the conceptual framework for 3D digital management of urban land and property information is proposed. The suggested framework comprises creating 3D models of urban land and property, validating the integrity of the models, integrating legal and physical information in the various 3D models, and analyzing the federated 3D model with the query. This approach will contribute to integrating silos of urban land and property information which can not only be utilized to manage the complex urban built environment and but also have the potential to reduce costs associated with data duplication.



Automatic Generation of LoD1 City Models and Building Segmentation from Single Aerial Orthographic Images using Conditional Generative Adversarial Networks

Lukas Beer

TU-Berlin, Deutschland

3D city models play an important role in multiple applications, but creating them still requires effort using various possible techniques. This paper proposes a new machine-learning-based framework for generating 3D city models. With the help of conditional Generative Adversarial Networks and single orthographic images, segmentation and height estimations of buildings are achieved. The height information per pixel and the building coordinates were generalized using a histogram for heights and the Douglas-Peucker algorithm. The framework was evaluated by using variations of the same dataset (for the city of Berlin) to show possible differences due to changes of the image size and representation of the heights. The evaluation reveals that it is possible to generate block models with a mean absolute height error of 5.53m per building, a mean absolute height error for the whole raster of 1.32m, and a Jaccard Index of 0.55 for the segmentation. While the proposed framework for generating LoD1 city models does not attain the accuracy of previous techniques, our work represents a step towards successfully using machine learning for the automatic generation of city models and building segmentation.



Creation of Nominal Asset Value-Based Map with GIS: Case Study of Istanbul Beyoğlu and Gaziosmanpaşa Districts

Muhammed Oguzhan Mete, Tahsin Yomralioglu

Istanbul Technical University, Turkey

The approach used in estimating the value of real estate has applications in fields as diverse as taxation, buying and renting properties, expropriation and urban regeneration. Determining the most objective, accurate and acceptable value for real estate by considering spatial criteria is important. One stochastic method used to determine real estate values is “nominal valuation”. In this approach, criteria that may affect land value are subjected to various spatial analyses and pixel-based value maps can be produced using GIS. Land value maps are in raster data format and need to be compared with the actual market values. Pixel-resolution analyses are required that depend on the selected grid dimensions. First of all, nominal value maps were produced using a nominal valuation model, using criteria for proximity, visibility and terrain. These were weighted in order to produce a nominal asset value-based map according to the “Best Worst Method”. Changes in the unit land values were examined for maps at various resolutions; a resolution of 10 meters emerged as the ideal pixel size for valuation maps.



 
Impressum · Kontaktadresse:
Datenschutzerklärung · Veranstaltung: GI_Week 2019
Conference Software - ConfTool Pro 2.6.129+TC
© 2001 - 2019 by Dr. H. Weinreich, Hamburg, Germany