University of Twente Student Theses


Exploiting shadows to infer 3D structures

Gadelmawla, Omar (2022) Exploiting shadows to infer 3D structures.

[img] PDF
Abstract:Buildings and vegetation height estimation using remotely sensed images are challenging to achieve. However, effective solutions to this challenge can serve in tackling more complex problems in the remote sensing field that require 3D information about objects in aerial images, which might be costly or inaccessible. Because shadows are a standard metric among many architectural building designs worldwide, shadows can help infer 3D structures as an auxiliary input in Deep Learning (DL) models. This paper proposes a method to combine RGB aerial imagery and height maps extracted from Light Detection And Ranging (LiDAR) sensors to develop an effective algorithm to realistically exaggerate shadow areas in RGB aerial imagery to enhance the learning process of the DL model. The results suggest that the proposed method is an effective solution for the problem, given the evaluation metrics specified in the paper.
Item Type:Essay (Bachelor)
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:54 computer science
Programme:Computer Science BSc (56964)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page