While the spectral information contained in hyperspectral images is rich, the spatial resolution of such images is in many cases very low. Many pixel spectra are mixtures of pure materials' spectra and therefore need to be decomposed into their constituents. This work investigates new decomposition methods taking into account spectral, spatial and global 3D adjacency information. This allows for faster and more accurate decomposition results.
Deep learning is widely applied to sparse 3D data to perform challenging tasks, e.g., 3D object detection and semantic segmentation. However, the high performance of deep learning comes with high costs, including computational costs and the effort to capture and label data. This work investigates and improves the efficiency of deep learning for sparse 3D data to overcome the obstacles to the further development of this technology.
Optical measurement methods are becoming increasingly important for high-precision production of components and quality assurance. The increasing demand can be met by modern imaging systems with advanced optics, such as light field cameras. This work explores their use in the deflectometric measurement of specular surfaces. It presents improvements in phase unwrapping and calibration techniques, enabling high surface reconstruction accuracies using only a single monocular light field camera.
While the spectral information contained in hyperspectral images is rich, the spatial resolution of such images is in many cases very low. Many pixel spectra are mixtures of pure materials' spectra and therefore need to be decomposed into their constituents. This work investigates new decomposition methods taking into account spectral, spatial and global 3D adjacency information. This allows for faster and more accurate decomposition results. This work was published by Saint Philip Street Press pursuant to a Creative Commons license permitting commercial use. All rights not granted by the work's license are retained by the author or authors.
In dieser Arbeit werden spektral kodierte multispektrale Lichtfelder untersucht, wie sie von einer Lichtfeldkamera mit einem spektral kodierten Mikrolinsenarray aufgenommen werden. Für die Rekonstruktion der kodierten Lichtfelder werden zwei Methoden entwickelt, eine basierend auf den Prinzipien des Compressed Sensing sowie eine Deep Learning Methode. Anhand neuartiger synthetischer und realer Datensätze werden die vorgeschlagenen Rekonstruktionsansätze im Detail evaluiert. -In this work, spatio-spectrally coded multispectral light fields, as taken by a light field camera with a spectrally coded microlens array, are investigated. For the reconstruction of the coded light fields, two methods, one based on the principles of compressed sensing and one deep learning approach, are developed. Using novel synthetic as well as a real-world datasets, the proposed reconstruction approaches are evaluated in detail.
In this work, the task of wide-area indoor people detection in a network of depth sensors is examined. In particular, we investigate how the redundant and complementary multi-view information, including the temporal context, can be jointly leveraged to improve the detection performance. We recast the problem of multi-view people detection in overlapping depth images as an inverse problem and present a generative probabilistic framework to jointly exploit the temporal multi-view image evidence.
This work presents model-based algorithmic approaches for interference-invariant time delay estimation, which are specifically suited for the estimation of small time delay differences with a necessary resolution well below the sampling time. Therefore, the methods can be applied particularly well for transit-time ultrasonic flow measurements, since the problem of interfering signals is especially prominent in this application.
The increasing use of automated laser welding processes causes high demands on process monitoring. This work demonstrates methods that use a camera mounted on the focussing optics to perform pre-, in-, and post-process monitoring of welding processes. The implementation uses machine learning methods. All algorithms consider the integration into industrial processes. These challenges include a small database, limited industrial manufacturing inference hardware, and user acceptance.
This book reviews the state of the art in algorithmic approaches addressing the practical challenges that arise with hyperspectral image analysis tasks, with a focus on emerging trends in machine learning and image processing/understanding. It presents advances in deep learning, multiple instance learning, sparse representation based learning, low-dimensional manifold models, anomalous change detection, target recognition, sensor fusion and super-resolution for robust multispectral and hyperspectral image understanding. It presents research from leading international experts who have made foundational contributions in these areas. The book covers a diverse array of applications of multispectral/hyperspectral imagery in the context of these algorithms, including remote sensing, face recognition and biomedicine. This book would be particularly beneficial to graduate students and researchers who are taking advanced courses in (or are working in) the areas of image analysis, machine learning and remote sensing with multi-channel optical imagery. Researchers and professionals in academia and industry working in areas such as electrical engineering, civil and environmental engineering, geosciences and biomedical image processing, who work with multi-channel optical data will find this book useful.
Resolving Spectral Mixtures: With Applications from Ultrafast Time-Resolved Spectroscopy to Superresolution Imaging offers a comprehensive look into the most important models and frameworks essential to resolving the spectral unmixing problem—from multivariate curve resolution and multi-way analysis to Bayesian positive source separation and nonlinear unmixing. Unravelling total spectral data into the contributions from individual unknown components with limited prior information is a complex problem that has attracted continuous interest for almost four decades. Spectral unmixing is a topic of interest in statistics, chemometrics, signal processing, and image analysis. For decades, researchers from these fields were often unaware of the work in other disciplines due to their different scientific and technical backgrounds and interest in different objects or samples. This led to the development of quite different approaches to solving the same problem. This multi-authored book will bridge the gap between disciplines with contributions from a number of well-known and strongly active chemometric and signal processing research groups. Among chemists, multivariate curve resolution methods are preferred to extract information about the nature, amount, and location in time (process) and space (imaging and microscopy) of chemical constituents in complex samples. In signal processing, assumptions are usually around statistical independence of the extracted components. However, the chapters include the complexity of the spectral data to be unmixed as well as dimensionality and size of the data sets. Advanced spectroscopy is the key thread linking the different chapters. Applications cover a large part of the electromagnetic spectrum. Time-resolution ranges from femtosecond to second in process spectroscopy and spatial resolution covers the submicronic to macroscopic scale in hyperspectral imaging. Demonstrates how and why data analysis, signal processing, and chemometrics are essential to the spectral unmixing problem Guides the reader through the fundamentals and details of the different methods Presents extensive plots, graphical representations, and illustrations to help readers understand the features of different techniques and to interpret results Bridges the gap between disciplines with contributions from a number of well-known and highly active chemometric and signal processing research groups
