3D LIDAR data application 3D LIDAR data application for urban morphogenesis for urban morphogenesis multi-agent vector based multi-agent vector based geosimulation geosimulation Vitor Silva, Dr. Corinne Plazanet, Cláudio Carneiro, Pr. François Golay 2008 ICCSA conference GEOG-AN-MOD session 2008 July the 2 nd
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
3D LIDAR data application 3D LIDAR data application
for urban morphogenesis for urban morphogenesis
multi-agent vector based geosimulationmulti-agent vector based geosimulation
Vitor Silva, Dr. Corinne Plazanet, Cláudio Carneiro, Pr. François Golay
2008 ICCSA conferenceGEOG-AN-MOD session
2008 July the 2nd
2
Context of the research Mega cities dynamics = complex auto-organised systems
Goal = develop simulation platform for decision support
=> assess scenarios of impacts of new architectural programs
Features (cadastre + SwissTopo datas): Buildings + their programmatic use Built environment : roads, railway, bus Natural environment: rivers, lake, green areas Administrative limits
+ LIDAR data => construct accurate 3D urban surface model for visibility analysis
Cognitive agents buildings equipments public places residential groups urban systems Towns
Objects of the environment Roads Lakes Rivers Green areas Railways and stations Public transports
BUILDINGS
Residential Equipment
PUBLIC PLACES
Street Place Park
MICRO
MESO
MACRO
RESIDENTIAL GROUP URBAN SYSTEM
TOWN
MEGA CITY
4
Laws Agents have different behaviours according to programmatic use
and environment
5 kinds of laws : Growing: probabilistic residential and derived equipment growing rate Stability: probabilistic end of life according to predefined thresholds Influence: programmatic influence between programs (services /
Visibility on particular attractions (e.g. lake Geneva)
Sunshine exposure
5
I am a new villa, I want to be close to school and have view to lake
School
SCHOOLLake
School
SCHOOLLake
Possible locations
Here !
6
Satisfaction degree of an agent
Visibility Satisfaction
Degree
For each agent a
n
ii
n
iii
solCoefvisCoefaaluenceCoef
solCoefsaDSvisCoefvaDSaaluenceCoefaaDSaDS
1
1
__),(inf_
_*),(_*),(),(inf_*),()(
[%]),( vaDS
Solar exposition satisfaction
degree
[%]),( saDS
For each influencing function f
If « asking » relation beetween a and f
Type_influence
If positive influence
If negative influence
Search for the closest
influencing agent/object ai
)),(_
)(),(1(*100),(
aiaInfluenceMax
aiaiadaiaDS
If « carrying » relation beetween a and f
Search for all influencing
agent/object aij
n
j ij
ijij
aaInfluenceMax
aaad
naiaDS
1
)),(_
)(),(1(100
1),(
Search for all influencing
agent/object aij
)),(_
)(),((*100),(
aiaInfluenceMax
aiaiadaiaDS
DS = Weighted average of satisfaction criteria
7
3D urban surface model
Construction of 3D urban surface modelLIDAR data, 6-8 pts / m2
Vertical resolution: 15 cm, Horizontal resolution: 20 cm
Normalised data height = terrain – building elevation
Cadastre (vectorial building footprints)
D.T.M.
Building model
LIDAR roof coverage
good insufficient
8
Visibility analysis: 2 perspectives
Using directly 3D urban surface model, from top and ground surfaces
Over buildings façades, in a vertical direction (elevation)
Grid T(i,j)Resolution: 1m/1m
Visibility = ∑ i,j T(i,j)
Top d
ow
n
9
Example of visibility analysis~
12 k
ms
~ 2,5 kms
Constrained to narrow angle, due to heavy time consuming computation
For each point, 2-5 minutes computation(3’000’000 pixels)
Pilot area: 500 × 500 meters
10
Results
Simulation without visibility parameter
Simulation with visibility parameter
Lake Geneva
Lake GenevaLake Geneva
11
Conclusions Conception of the approach, first results
Visibility indicators have great importance for localization => precision of LIDAR data particularly interesting
Improvements on the simulations results => better interpretation of importance of LIDAR data
Time consuming
Great perspectives …
Implement the approach into the system would allow: Integrate visibilty and sunshine analysis at each transition Dynamic visualisation Compute potential visibility for new buildings