View
5
Download
0
Category
Preview:
Citation preview
RSCy2017March 20 – 23, 2017 - Cyprus
GEOSPATIAL DATA AND GEOINFORMATICS TECHNOLOGIES SUPPORTING SMART CITY STRATEGIES –
THE SENECA PROJECTE. Nocerino a, I. Toschi a, F. Remondino a
A. Revolti b, G. Soria b, S. Piffer b
a 3D Optical Metrology (3DOM), Bruno Kessler Foundation (FBK), Trento, Italy
b Trilogis Srl, Trento, Italy
“Smart cities”: future trend
• According to the UN World Urbanization Trend (http://esa.un.org/), in 2050 about66% of the world population will live in urban areas.
• By 2030, the world is projected to have 41 mega-cities with more than 10 millioninhabitants.
• It is clear that a more efficient mapping, understanding and management of theurban ecosystem is required → “Smart City” concept
All these numbers and future trend encourage the rapidly growing sector of
3D GEOSPATIAL DATA AND GEOINFORMATICS TECHNOLOGIES
SUPPORTING SMART CITY STRATEGIES
www.eureka-smart-cities.org
“Smart cities”: 3D City Modelling
The need for a realistic 3D modelling for cities is evident and an increasing spectrum of potential applications urgently demands advanced methods
for efficient and automated urban reconstruction.
3D City Modelling: open issues
The potential of 3D data is mainly confined to visualization purposes, although the geometry and the appearance of 3D building models are just the tip of the iceberg.
Significant manual editing and unrealistic assumptions are still required, although the geometry of historic city centres is far away from the “Manhattan-world”.
Oblique airborne photogrammetry is rapidly maturing, although its potential is still underexploited especially when it comes to the use of oblique views for façade modelling.
The SENECA project
Therefore, the major challenge today is to create automatic procedures that make best use of available technologies and data.
Geospatial data modelling
Geoinformatics technologies development
3DOM = 3D Optical Metrology
Bruno Kessler Foundation
High-tech SME for innovative GIS solutions
Smart and SustaiNablE City from Above
seneca.fbk.eu
The SENECA project
In this framework, the project aims to support the actual exploitation and valorisation of urban 3D reconstruction by developing innovative solutions that
GEOMETRY CAPTURE: 3D POINT CLOUD
GEOMETRY MODELLING: 3D BUILDING MODELS
GEOMETRY ENRICHMENT:3D GEO-DATABASE
MANAGEMENT SYSTEM
(i) efficiently exploit the aerial photogrammetric workflow (aerial triangulation and dense image matching)
(ii) derive topologically and geometrically accurate 3D geo-objects with a multi-scale approach
(ii) link geometries with ancillary information within a scalable and expandable 3D environment
Case studies
The developed methodology is tested on two test sites, that vary in urban characteristics and pose different reconstruction challenges.
Google Earth
Trento (Italy)
Population: 117.000 inhabitantsTest area: 3.5 x 1.5 kmGeography: it lies on the banks of the river Adige, in the homonymous valley. It is surrounded by the Alps.Urban characteristics: it presents older buildings with complex shapes (Romanesque, Medieval and Renaissance style), located in a densely built city centre. The outskirts contains residential and commercial buildings.
Graz (Austria)
Population: 280.000 inhabitantsTest area: 3.0 x 1.5 km
Geography: it is situated on the banks of the river Mur. It lies in a basin, southeast of
the Alps.Urban characteristics: it presents a
medieval main square surrounded by narrow streets, with Renaissance and Baroque architectures. The outskirts
features single houses in a residential area.
Input: spatial data
1. VIS airborne oblique and/or nadir images
UltraCamX Prime (Trento – flown by AVT)
UltraCam Osprey I(Graz – flown by Vexcel)
Nadir Nadir Oblique 45°
Sensor size (mm) 103.86 x 67.86 70 x 45 23.5 x 36 (L/R)
71.5 x 23.5 (F/B)
Focal length (mm) 100.5 51 80
GSD (m) ~0.10 ~0.12(*)
Y/X Overlap (%) 80/60 75/65(*)
# Images Nadir/Oblique 397 20/160
# GCP/CP 14/6 4/3
(*) Computed on nadir images
Source: Vexcel Imaging
Input: spatial data
2. Thermal orthophoto
0.0 km 3.4 km
10
5
0
-5
-10
-15
-20
[C°]
TABI-1800(Graz– flown by AVT)
Date and time20 Dec. 2011
19:38 – 22:02 CET
Processing
• Raw thermal data processed by ITRES Research
• Orthorectification based on LiDAR DSM
Orthophoto spatial resolution [m]
0.60
Source: AVT and ITRES
3. Land cadastre map with building footprints (and land parcel IDs)
Input: non-spatial data
3. Energy Performance Certificates incl. • energy consumptions• carbon dioxide emissions• energy efficiency indexes• etc.
4. Sources of artificial light (single spots)• streetlights• advertising• etc.
5. Data from the register of buildings• owner• number of rooms• surface• property category• etc.
Collaboration with local public
authorities, SMES and private
organizations
Geometry capture: 3D point cloud
Trento Graz
TrentoAT (Nadir images, 6 CPs)
RMSE X [m] 0.08
RMSE Y [m] 0.05
RMSE Z [m] 0.09
GNSS and GCPs observations includedin the BBA as observed unknowns
2.5D DSM CLOUD5.5 billion points,
10 cm spatial resolution
Geometry capture: 3D point cloud
Trento
Geometry capture: 3D point cloud
Graz AT (Nadir and Oblique images, 3 CPs)
RMSE X [m] 0.05
RMSE Y [m] 0.04
RMSE Z [m] 0.02 (0.08 only N)
• The BBA is supported by the calibrated relative orientation parameters
3D POINT CLOUD60 million points
(subset), 10 cm spatial
resolution
• GNSS and GCPs observations included in the BBA as observed unknowns
Geometry capture: 3D point cloud
Graz
Geometry capture: 3D point cloud
Geometry modelling: 3D building models
Trento
Graz
Graz
(*) According to the OGC standard CityGML
(*)
Geometry modelling: 3D building models
Trento At urban scale: LOD1 building models• polyhedral models;• input: Digital Terrain Model and cadastral
footprints;• Footprints are extruded to a uniform average
building height.
At neighbourhood scale: LOD2 building models• models composed by complex roofs and vertical
walls connecting them;• input: DMS cloud, Digital Terrain Model and
cadastral footprints;• segmentation process: the result is a partition of
the points, where all points in one segment belong to the same shape.
• building generation: the detection of building objects is performed by searching for selected types of internal- (or user-) defined roof types.
At building scale: LOD3 building models• architectural models with detailed wall and roof
structures;• input: cadastral footprints and data from the
register of buildings;• building models are generated using CAAD-based tools
Geometry modelling: 3D building models
Graz 10
5
0
-5
-10
-15
-20
[C°]
3D MESHtextured with VIS images
3D MESHtextured with THERMAL ortho
Work in progress:• orthophoto classification;• semantic labelling transfer from 2D to 3D
data;• urban geometric reconstruction
supported by semantic knowledge• semantically-enriched 3D city modelling
BUILDING GRASS TREES STREET
GEOMETRY ENRICHMENT: 3D GEO-DATABASE SYSTEM
Platform architecture
Data access layer:• it provides access to the DB, PostgreSQL with PostGIS, in-house customized • DB stores geometry at different LODs, each building is associated to a unique ID;• queries are performed on spatial data to retrieve ancillary info, and vice versa.
Raw data management layer:• it allows to manipulate raw geo-data and import them in the data access layer.
Business logic layer:• it allows to manage permissions and filters for a proper handle of sensitive and
confidential data.
Service layer:• it manages data exchange with the other layers and provides public APIs for
data import and integration with third-party sw.
Presentation layer:• it is the web application that makes data available to the end-users.• it manages 3D view of geometries, texture projection and render of map layers. • 3D navigation and rendering are performed using the virtual globe Web World
Wind (Nasa), in-house customized.
Multi-scale navigation and queryFrom LOD1 models at urban scale, to LOD2
photo-textured models at neighborhood scale
From LOD2 models, to LOD3architectural models with
associated ancillary information
GEOMETRY ENRICHMENT: 3D GEO-DATABASE SYSTEM
Conclusions and future works
Future works within the SENECA framework will be aimed to:
Extend the platform content
• open data issued by public administrations (harmonisation strongly required);• further environmental pollution sources (e.g. CO2 emission and noise pollution);• direct links to publicly-available DBs.
Further support the “Smart” exploitation of the system
• improvement of the multi-scale rendering performance;• space management in “true” 3D space;• simulations (energy heating demand, PV-suitability of the roofs, noise pollution).
Encourage an active involvement of the public
• from “nice-to-have” to “need-to-have”;• organization of public events to show results and receive feedbacks;• create dialogue among potential users (PA, real estate companies, security
companies, private owners, general public, etc.).
Conclusions and future works
The need for a realistic 3D modelling for cities is evident and an increasing spectrum of potential applications urgently demands advanced methods
for efficient and automated urban reconstruction.
SENECA demonstrated a step forward
towards:
The actual exploitation of 3D data potential:• 3D city modelling is far more then the 3D realistic
visualization of urban environment;• semantic modelling is the future key issue.
The urban reconstruction of complex scenarios:• 3D modelling of historic city centers requires an
improvement of traditional reconstruction methods;• cadastral registration needs to progress to a 3D
approach.
The use of oblique views to support 3D city modelling:• the inclusion of obliques asks for an adjustment of
traditional photogrammetric pipeline;• slanted views can provide detailed info on building facades,
however integration with terrestrial data is still required.
GEOSPATIAL DATA AND GEOINFORMATICS TECHNOLOGIES SUPPORTING SMART CITY STRATEGIES –
THE SENECA PROJECT
Isabella ToschiFBK - Fondazione Bruno Kessler
3DOM – 3D Optical Metrology UnitTrento, Italy
web: http://3dom.fbk.eu email: toschi@fbk.eu
http://seneca.fbk.eu/
RSCy2017March 20 – 23, 2017 - Cyprus
Recommended