Vidushi Bhatt
Voxel based 3D Point Cloud Semantic Segmentation at Urban scale

Most 3D PC segmentation techniques today either attempt ML methodologies to achieve segmentation or labour intensive classification by people. The issue with ML techniques for segmentation is primarily related to the lack of sufficient training material to execute achieve a reliable classification. New research in this direction experiments with the utilisation of voxel structures to attempt segmentation of point cloud. This project is aimed at looking into the current techniques and assessing their applicability at the urban scale.

Supervisors: Peter Van Oosterom + Undecided

Aswathy Chandran
Building Material component for the LADM

The thesis focuses on explores the potential integration of Building Material into the Land Administration Domain Model (LADM). It centers on advancing the circularity and sustainability of buildings by introducing an additional component in LADM. This element will consists of a database of materials organized by their components like quality, type etc, offering an understanding of their estimated lifespan. These materials can then be reused in new construction projects, taking into account geographical considerations like proximity and the type of building involved. The research aims to evaluate the feasibility and advantages of incorporating the building materials section into the existing LADM framework, utilizing case studies. Through this project, it aims to understand the implications for improving the circularity and efficiency of both land administration systems and building materials.

Supervisors: Peter van Oosterom + Wilko Quak

Susanne Epema
A 3D delineation method for urban river spaces.

An accurate spatial description, a spatial delineation, of riverside urban areas is essential for a wide range of applications, including flood mitigation and public space design. This thesis aims to develop a 3D delineation method for urban river spaces by adapting an existing 2D delineation approach. By adapting this 2D method, the aim is to develop a technique that is able to more accurately represent the spatial complexities of these areas. This thesis will tackle challenges related to using open 3D data effectively and exploring scalability options within the Netherlands but also possibly internationally.

Supervisors: Claudiu Forgaci + Jantien Stoter

Eirini Chrysovalantou Tsipa
Usability Evaluation of Vario-Scale Maps for Dynamic and Interactive Mapping

The digital mapping environment is evolving from traditional fixed-scale maps to dynamic, interactive solutions. This thesis proposes a usability study of the ‘vario-scale’ approach, aimed at providing smoother map interactions.

Objective The main objective is to assess the usability of vario-scale maps and suggest empirical enhancements. The study will evaluate the user’s ability to perform realistic map-use tasks, focusing on ease of use, user satisfaction, and interaction effectiveness.

Methodology The first step in this study will be a comprehensive literature review. Next, prototypes will be developed following cartographic principles. The study will explore the impact of map properties like label placement, and vario-scale-specific properties, such as feature importance. The study participants will perform map-use tasks, vocalising their thoughts to capture interaction impressions and cognitive processes. For this, eye-tracking and screen-logging software will be utilised.

Challenges A significant challenge is user adaptability, as traditional map users might find the vario-scale concept initially confusing. Additionally, the study will require assessing the participants’ backgrounds to tailor usability tests, accommodating various familiarity levels. Developing targeted prototypes that gradually introduce vario-scale features will be challenging. Identifying optimal modifications of the map’s properties for each prototype is crucial.

Supervisors: Martijn Meijers + Peter Van Oosterom

Longxiang Xu
Large-scale solar potential analyses based on Semantic 3D City Models

Solar potential analysis is of great help to energy transition through city planning, analysis of occupant behavior, assessing building energy performance, and scheduling the power grid. Computing high resolution (spatial, temporal, and semantic) solar irradiance is the key step of solar potential analysis. To determine the solar irradiance with high resolution, we can utilize 3D City Models (3DCM). Within the context of 3DCM, calculating shadows, sky-view factor (SVF), and ground view factor (GVF) are the most important and computationally intensive steps. Existing methods for the calculations, either apply simplifications or assumptions to make the computation achievable in a limited time. While some methods are accurate, but unable to scale to city-level. The thesis will be focused on developing a methodology that can achieve fast shadow solar irradiance with minimum assumptions. The expected methodology will be able to be applied to large areas such as the whole Netherlands.

Supervisors: Camilo León Sánchez + Giorgio Agugiaro