A Domain Decompositionbased CNN Architecture for High-Resolution Image Segmentation

This thesis addresses the challenge of segmenting ultra-high-resolution images. Limitations of current approaches to segment these are that either detailed spatial contextual information is lost or many redundant computations are necessary. To overcome these issues, we propose a novel approach combining the U-Net architecture with domain...

Deep Learning-based Segmentation of Cracks within a Photogrammetry Solution: Fully-Supervised Learning, Transfer Learning and Photogrammetric Image Processing

The city of Amsterdam faces the challenge of monitoring and assessing 200 kilometers of historic quay walls, of which much is deemed to be in poor condition. A key monitoring technique used is photogrammetry resulting in deformation testing. The fundamental data source forming the basis of this deformation analysis is a collection of overlapping...

Quantitative Analysis of Flame Kinematics in Premixed Hydrogen-Air-DNG Jet Flames using PIV Measurements and Flame Front Segmentation

Today's electricity supply falls short of current demands, leading to the utilization of gas turbines in both ground based and avionic infrastructures. Nevertheless, these often rely on carbon-based fuels, resulting in escalating CO2 emissions. However, adopting hydrogen as a fuel eliminates carbon emissions. Aside from zero carbon emissions,...

Leverage Spatial Priors and Copy-Paste to Boost Contrastive Learning for Urban-Scene Segmentation

Self-supervised contrastive learning has achieved remarkable performance in computer vision. Its success relies on certain priors that vary from different tasks and data at hand, e.g, the object-centric prior implied by ImageNet. For segmentation on complex scenes, researchers have introduced salient objects or auxiliary labels as priors to...

Neural Architecture Search for Medical Image Segmentation

Deep Neural Networks (DNNs) have the potential to make various clinical procedures more time-efficient by automating medical image segmentation; largely due to their strong, in some cases human-level, performance. The design of the best possible medical image segmentation DNN, however, is task-specific. Neural Architecture Search (NAS), i.e.,...

Automating malaria diagnosis: a machine learning approach: Erythrocyte segmentation and parasite identification in thin blood smear microscopy images using convolutional neural networks

Reliable malaria diagnosis techniques that are suitable for point-of-care testing in high burden areas, are vital for effective treatment and monitoring of the disease. Identification of malaria parasites in Giemsa stained blood slides is currently the most widely accepted technique, but its availability is limited by the need for highly trained...

Landslide Detection using Random Forest Classifier

Landslides are destructive and recurrent natural disasters that cost annually significant social and economic losses all over the world. These events can be induced by natural factors as earthquakes and extreme rainfall, as well as by human intervention, including construction and mining. A primary resource to conduct landslides studies for...

Image Segmentation of the γ'-Phase in Nickelbase Superalloys utilising Deep Learning

Quantitative analysis of material microstructure is a well-known method to derive chemical and physical properties of a sample. This includes the segmentation of e.g. Light Optical Microscopy or Scanning Electron Microscopy images where each pixel is assigned to a material. Since some phases such as the γ-γ’ structure in nickelbased superalloys...

Segmentation of Carpal Bones: Based on Statistical Shape Models using Spherical Harmonics

A study has been conducted on the application of Spherical Harmonics based Statistical Shape Modelling for Image Segmentation. This study is focused on the segmentation of Wrist bones using the above mentioned technique.

Moving object detection and image inpainting in street-view imagery

In this thesis, a pipeline is created consisting of two parts. In the first part, the moving objects (cars, cyclists, pedestrians) are detected in street-view imagery using image segmentation neural networks and a LIDAR-based moving object detection approach. In the second part, those moving objects are deleted from the image data and an image...

The effect of mechanical stratigraphy on fracture pattern evolution and permeability in faulted carbonates - insights from analogue sandbox models

Dilatant fractures and faults in carbonate rocks play an important role as fluid path for many applications. However, multidimensional characterization of these fractures remains a challenge due to structural complexity in multiscale. The progressive deformation of dilatant fracture in normal fault system from physical models are imaged in 3D...

Image segmentation assisted by a triple-camera setup

Photographers suffer from many challenging problems when separating and compositing images, such as the color camouflage and unpleasant color bleeding artifacts. To tackle these issues, we present a solution to fore- and background segmentation for images taken from multiple cameras, which introduce the parallax that leads to the foreground...

Deep Learning for Pixelwise Classification of Hyperspectral Images: A generalizing model for a fixed scene subject to temporally changing weather, lighting and seasonal conditions

In hyperspectral (HS) imaging, for every pixel a spectrum of wavelengths is captured. These spectra represent material properties, i.e. the spectral signatures. So, classification of HS imagery is based on material properties. This thesis describes a framework to perform pixelwise classification of HS images of a fixed scene subject to varying...

Learn representations in the presence of segmentation label noises

Training data for segmentation tasks are often available only on a small scale. Transferring learned representations from pre-trained classification models is therefore widely adopted by convolutional neural networks for semantic segmentation. In domains where the representations from the classification models are not directly applicable, we...

MR Scanner Invariant Feature Representation using Label Transfer and Clustering

Machine learning approaches are increasingly successful in medical image analysis. Still, learning from MR images poses some serious challenges. Scanner-dependent characteristics effect feature representations directly and hamper the clinical implementation of otherwise successful supervised-learning techniques. To compensate for variations in...

Scalable machine learning algorithms on a big data infrastructure

Two currently popular topics in computer science are machine learning and big data. Often the two are combined to obtain powerful machines with learning capabilities or high throughput data analysis programs among others. This research analyses which machine learning techniques qualify to be efficiently implemented on a scalable big data...

Mapping Urban Surface Infiltration Capacity: Segment-based land cover classification with VHR imagery for urban water management and design.

Effective urban water management requires certainty about surface conditions such as the surface infiltration capacity (SIC). A SIC map of an urban catchment area could be a useful input for evidence-based urban design of water-sensitive/low-impact neighbourhoods and multitiered water management schemes for reducing flood vulnerability. Methods...