Skip to main content

LUP Student Papers

LUND UNIVERSITY LIBRARIES

Parametric Roof Modelling Using a Machine Learning Approach

Åberg, Iris LU and Åsvärd, Thelma LU (2026) In Thesis in geographical information technics EXTM05 20261
Department of Earth and Environmental Sciences (MGeo)
Dept of Physical Geography and Ecosystem Science
Abstract
Three-dimensional building models are becoming increasingly important in areas such as urban planning, simulations, and geographical analyses. At the same time, detailed information about roof structures is often missing from current geospatial datasets, which complicates the automated reconstruction of coherent and functional 3D building models. Point cloud data acquired through laser scanning enables detailed geometric representation of buildings, but classification of roof structures from point clouds remains a complex and resource-demanding task. In addition, the application of machine learning methods is limited by the lack of annotated training data.
This thesis investigates the possibility of classifying real-world roof structures... (More)
Three-dimensional building models are becoming increasingly important in areas such as urban planning, simulations, and geographical analyses. At the same time, detailed information about roof structures is often missing from current geospatial datasets, which complicates the automated reconstruction of coherent and functional 3D building models. Point cloud data acquired through laser scanning enables detailed geometric representation of buildings, but classification of roof structures from point clouds remains a complex and resource-demanding task. In addition, the application of machine learning methods is limited by the lack of annotated training data.
This thesis investigates the possibility of classifying real-world roof structures based on point cloud data using a machine learning model trained on synthetically generated data. The study also examines how classified roof types can be represented parametrically and used for reconstruction of simplified 3D building models.
The methodology consists of four main stages:
  • Generation of parametric 3D building models
  • Creation of synthetic point clouds
  • Training and evaluation of a machine learning model
  • Reconstruction of 3D buildings based on classified roof types.

Parametric building models are generated in CityEngine and FME using CGA rules, where parameters such as roof type, building height, and roof height are varied. The generated models are converted into synthetic point clouds, rasterized into image representations, and used to train a ResNet-based deep learning model. Real-world LiDAR data from Gothenburg were used for testing.
The results show that the model achieves high classification accuracy when both training and testing are performed on synthetic data. However, the accuracy decreases significantly when the model is applied to real-world LiDAR data, indicating that generalization between synthetic and real data remains a major challenge. The study also demonstrates that certain roof types are difficult to distinguish due to similar geometric characteristics, variations in real building structures, and differences between synthetic and real data. The classified roof types were subsequently linked to parametric roof descriptions and used to reconstruct simplified 3D building models.
In conclusion, the study demonstrates that synthetically generated training data combined with parametric modeling and machine learning has potential for automated roof classification and parametric representation of buildings from point cloud data. At the same time, improved generalization towards real-world data and more representative training datasets are identified as important areas for future development. (Less)
Abstract (Swedish)
Tredimensionella byggnadsmodeller blir allt viktigare inom områden såsom stadsplanering, simuleringar och geografiska analyser. Samtidigt saknas ofta detaljerad information om byggnaders takstrukturer i befintliga geodata, vilket försvårar automatiserad rekonstruktion av sammanhängande och användbara 3D-modeller. Punktmolnsdata från laserskanning möjliggör detaljerad geometrisk representation av byggnader, men klassificering av takstrukturer från punktmoln är fortfarande en komplex och resurskrävande process. Dessutom begränsas användningen av maskininlärning inom området av bristen på annoterade träningsdata.
Detta examensarbete syftar till att undersöka möjligheten att klassificera verkliga takstrukturer baserat på punktmolnsdata genom... (More)
Tredimensionella byggnadsmodeller blir allt viktigare inom områden såsom stadsplanering, simuleringar och geografiska analyser. Samtidigt saknas ofta detaljerad information om byggnaders takstrukturer i befintliga geodata, vilket försvårar automatiserad rekonstruktion av sammanhängande och användbara 3D-modeller. Punktmolnsdata från laserskanning möjliggör detaljerad geometrisk representation av byggnader, men klassificering av takstrukturer från punktmoln är fortfarande en komplex och resurskrävande process. Dessutom begränsas användningen av maskininlärning inom området av bristen på annoterade träningsdata.
Detta examensarbete syftar till att undersöka möjligheten att klassificera verkliga takstrukturer baserat på punktmolnsdata genom att använda en maskininlärningsmodell tränad på syntetiskt genererade data. Arbetet undersöker även hur taktyper kan beskrivas parametriskt och användas för rekonstruktion av förenklade 3D-byggnadsmodeller.
Metoden är uppbyggd kring fyra huvudsakliga steg:
  • Generering av parametriska 3D-byggnadsmodeller
  • Skapande av syntetiska punktmoln
  • Träning och utvärdering av en maskininlärningsmodell
  • Rekonstruktion av 3D-byggnader baserat på klassificerade taktyper.

Parametriska byggnadsmodeller genererades i CityEngine och FME med hjälp av CGA-regler där parametrar såsom taktyp, byggnadshöjd och takhöjd varierades. De genererade modellerna omvandlades därefter till syntetiska punktmoln som rasteriserades till bilder och användes för träning av en ResNet-baserad djupinlärningsmodell. För testning användes riktig LiDAR-data från Göteborg.
Resultaten visar att modellen uppnår hög klassificeringsnoggrannhet när både träning och testning utförs på syntetiska data. Däremot minskar noggrannheten tydligt när modellen tillämpas på riktig LiDAR-data, vilket visar att generalisering mellan syntetiska och verkliga data fortfarande är en stor utmaning. Studien visar även att vissa taktyper är svåra att skilja åt på grund av liknande geometriska egenskaper, variationer i verkliga byggnadsstrukturer och skillnader mellan syntetiska och verkliga data. De klassificerade taktyperna kunde därefter kopplas till parametriska takbeskrivningar och användas för att rekonstruera förenklade 3D-
byggnadsmodeller.
Sammanfattningsvis visar studien att syntetiskt genererade träningsdata i kombination med parametrisk modellering och maskininlärning har potential för automatiserad klassificering och parametrisk representation av byggnader från punktmolnsdata. Samtidigt framgår att förbättrad generalisering mot verkliga data och mer representativa träningsdata är centrala områden för fortsatt utveckling. (Less)
Please use this url to cite or link to this publication:
author
Åberg, Iris LU and Åsvärd, Thelma LU
supervisor
organization
course
EXTM05 20261
year
type
H2 - Master's Degree (Two Years)
subject
publication/series
Thesis in geographical information technics
report number
45
language
English
id
9239903
date added to LUP
2026-06-24 09:25:41
date last changed
2026-06-24 09:25:41
@misc{9239903,
  abstract     = {{Three-dimensional building models are becoming increasingly important in areas such as urban planning, simulations, and geographical analyses. At the same time, detailed information about roof structures is often missing from current geospatial datasets, which complicates the automated reconstruction of coherent and functional 3D building models. Point cloud data acquired through laser scanning enables detailed geometric representation of buildings, but classification of roof structures from point clouds remains a complex and resource-demanding task. In addition, the application of machine learning methods is limited by the lack of annotated training data.
This thesis investigates the possibility of classifying real-world roof structures based on point cloud data using a machine learning model trained on synthetically generated data. The study also examines how classified roof types can be represented parametrically and used for reconstruction of simplified 3D building models.
The methodology consists of four main stages:
[list]
[*]Generation of parametric 3D building models
[*]Creation of synthetic point clouds
[*]Training and evaluation of a machine learning model
[*]Reconstruction of 3D buildings based on classified roof types.
[/list]
Parametric building models are generated in CityEngine and FME using CGA rules, where parameters such as roof type, building height, and roof height are varied. The generated models are converted into synthetic point clouds, rasterized into image representations, and used to train a ResNet-based deep learning model. Real-world LiDAR data from Gothenburg were used for testing.
The results show that the model achieves high classification accuracy when both training and testing are performed on synthetic data. However, the accuracy decreases significantly when the model is applied to real-world LiDAR data, indicating that generalization between synthetic and real data remains a major challenge. The study also demonstrates that certain roof types are difficult to distinguish due to similar geometric characteristics, variations in real building structures, and differences between synthetic and real data. The classified roof types were subsequently linked to parametric roof descriptions and used to reconstruct simplified 3D building models.
In conclusion, the study demonstrates that synthetically generated training data combined with parametric modeling and machine learning has potential for automated roof classification and parametric representation of buildings from point cloud data. At the same time, improved generalization towards real-world data and more representative training datasets are identified as important areas for future development.}},
  author       = {{Åberg, Iris and Åsvärd, Thelma}},
  language     = {{eng}},
  note         = {{Student Paper}},
  series       = {{Thesis in geographical information technics}},
  title        = {{Parametric Roof Modelling Using a Machine Learning Approach}},
  year         = {{2026}},
}