Kaiserslautern - Fachbereich Informatik
Refine
Year of publication
Document Type
- Doctoral Thesis (235) (remove)
Has Fulltext
- yes (235)
Keywords
- Visualisierung (20)
- Visualization (8)
- Deep Learning (7)
- Computergraphik (5)
- Evaluation (4)
- Robotik (4)
- Artificial Intelligence (3)
- Bildverarbeitung (3)
- Geoinformationssystem (3)
- Machine Learning (3)
Faculty / Organisational entity
Medical cyber-physical systems (MCPS) emerged as an evolution of the relations between connected health systems, healthcare providers, and modern medical devices. Such systems combine independent medical devices at runtime in order to render new patient monitoring/control functionalities, such as physiological closed loops for controlling drug infusion or optimization of alarms. Despite the advances regarding alarm precision, healthcare providers still struggle with alarm flooding caused by the limited risk assessment models. Furthermore, these limitations also impose severe barriers on the adoption of automated supervision through autonomous actions, such as safety interlocks for avoiding overdosage. The literature has focused on the verification of safety parameters to assure the safety of treatment at runtime and thus optimize alarms and automated actions. Such solutions have relied on the definition of actuation ranges based on thresholds for a few monitored parameters. Given the very dynamic nature of the relevant context conditions (e.g., the patient’s condition, treatment details, system configurations, etc.), fixed thresholds are a weak means for assessing the current risk. This thesis presents an approach for enabling dynamic risk assessment for cooperative MCPS based on an adaptive Bayesian Networks (BN) model. The main aim of the approach is to support continuous runtime risk assessment of the current situation based on relevant context and system information. The presented approach comprises (i) a dynamic risk analysis constituent, which corresponds to the elicitation of relevant risk parameters, risk metric building, and risk metric management; and (ii) a runtime risk classification constituent, which aims to analyze the current situation risk, establish risk classes, and identify and deploy mitigation measures. The proposed approach was evaluated and its feasibility proved by means of simulated experiments guided by an international team of medical experts with a focus on the requirements of efficacy, efficiency, and availability of patient treatment.
A prime motivation for using XML to directly represent pieces of information is the ability of supporting ad-hoc or 'schema-later' settings. In such scenarios, modeling data under loose data constraints is essential. Of course, the flexibility of XML comes at a price: the absence of a rigid, regular, and homogeneous structure makes many aspects of data management more challenging. Such malleable data formats can also lead to severe information quality problems, because the risk of storing inconsistent and incorrect data is greatly increased. A prominent example of such problems is the appearance of the so-called fuzzy duplicates, i.e., multiple and non-identical representations of a real-world entity. Similarity joins correlating XML document fragments that are similar can be used as core operators to support the identification of fuzzy duplicates. However, similarity assessment is especially difficult on XML datasets because structure, besides textual information, may exhibit variations in document fragments representing the same real-world entity. Moreover, similarity computation is substantially more expensive for tree-structured objects and, thus, is a serious performance concern. This thesis describes the design and implementation of an effective, flexible, and high-performance XML-based similarity join framework. As main contributions, we present novel structure-conscious similarity functions for XML trees - either considering XML structure in isolation or combined with textual information -, mechanisms to support the selection of relevant information from XML trees and organization of this information into a suitable format for similarity calculation, and efficient algorithms for large-scale identification of similar, set-represented objects. Finally, we validate the applicability of our techniques by integrating our framework into a native XML database management system; in this context we address several issues around the integration of similarity operations into traditional database architectures.
This thesis presents a novel, generic framework for information segmentation in document images.
A document image contains different types of information, for instance, text (machine printed/handwritten), graphics, signatures, and stamps.
It is necessary to segment information in documents so that to process such segmented information only when required in automatic document processing workflows.
The main contribution of this thesis is the conceptualization and implementation of an information segmentation framework that is based on part-based features.
The generic nature of the presented framework makes it applicable to a variety of documents (technical drawings, magazines, administrative, scientific, and academic documents) digitized using different methods (scanners, RGB cameras, and hyper-spectral imaging (HSI) devices).
A highlight of the presented framework is that it does not require large training sets, rather a few training samples (for instance, four pages) lead to high performance, i.e., better than previously existing methods.
In addition, the presented framework is simple and can be adapted quickly to new problem domains.
This thesis is divided into three major parts on the basis of document digitization method (scanned, hyper-spectral imaging, and camera captured) used.
In the area of scanned document images, three specific contributions have been realized.
The first of them is in the domain of signature segmentation in administrative documents.
In some workflows, it is very important to check the document authenticity before processing the actual content.
This can be done based on the available seal of authenticity, e.g., signatures.
However, signature verification systems expect pre-segmented signature image, while signatures are usually a part of document.
To use signature verification systems on document images, it is necessary to first segment signatures in documents.
This thesis shows that the presented framework can be used to segment signatures in administrative documents.
The system based on the presented framework is tested on a publicly available dataset where it outperforms the state-of-the-art methods and successfully segmented all signatures, while less than half of the found signatures are false positives.
This shows that it can be applied for practical use.
The second contribution in the area of scanned document images is segmentation of stamps in administrative documents.
A stamp also serves as a seal for documents authenticity.
However, the location of stamp on the document can be more arbitrary than a signature depending on the person sealing the document.
This thesis shows that a system based on our generic framework is able to extract stamps of any arbitrary shape and color.
The evaluation of the presented system on a publicly available dataset shows that it is also able to segment black stamps (that were not addressed in the past) with a recall and precision of 83% and 73%, respectively.
%Furthermore, to segment colored stamps, this thesis presents a novel feature set which is based on intensity gradient, is able to extract unseen, colored, arbitrary shaped, textual as well as graphical stamps, and outperforms the state-of-the-art methods.
The third contribution in the scanned document images is in the domain of information segmentation in technical drawings (architectural floorplans, maps, circuit diagrams, etc.) containing usually a large amount of graphics and comparatively less textual components. Further, as in technical drawings, text is overlapping with graphics.
Thus, automatic analysis of technical drawings uses text/graphics segmentation as a pre-processing step.
This thesis presents a method based on our generic information segmentation framework that is able to detect the text, which is touching graphical components in architectural floorplans and maps.
Evaluation of the method on a publicly available dataset of architectural floorplans shows that it is able to extract almost all touching text components with precision and recall of 71% and 95%, respectively.
This means that almost all of the touching text components are successfully extracted.
In the area of hyper-spectral document images, two contributions have been realized.
Unlike normal three channels RGB images, hyper-spectral images usually have multiple channels that range from ultraviolet to infrared regions including the visible region.
First, this thesis presents a novel automatic method for signature segmentation from hyper-spectral document images (240 spectral bands between 400 - 900 nm).
The presented method is based on a part-based key point detection technique, which does not use any structural information, but relies only on the spectral response of the document regardless of ink color and intensity.
The presented method is capable of segmenting (overlapping and non-overlapping) signatures from varying backgrounds like, printed text, tables, stamps, logos, etc.
Importantly, the presented method can extract signature pixels and not just the bounding boxes.
This is substantial when signatures are overlapping with text and/or other objects in image. Second, this thesis presents a new dataset comprising of 300 documents scanned using a high-resolution hyper-spectral scanner. Evaluation of the presented signature segmentation method on this hyper-spectral dataset shows that it is able to extract signature pixels with the precision and recall of 100% and 79%, respectively.
Further contributions have been made in the area of camera captured document images. A major problem in the development of Optical Character Recognition (OCR) systems for camera captured document images is the lack of labeled camera captured document images datasets. In the first place, this thesis presents a novel, generic, method for automatic ground truth generation/labeling of document images. The presented method builds large-scale (i.e., millions of images) datasets of labeled camera captured / scanned documents without any human intervention. The method is generic and can be used for automatic ground truth generation of (scanned and/or camera captured) documents in any language, e.g., English, Russian, Arabic, Urdu. The evaluation of the presented method, on two different datasets in English and Russian, shows that 99.98% of the images are correctly labeled in every case.
Another important contribution in the area of camera captured document images is the compilation of a large dataset comprising 1 million word images (10 million character images), captured in a real camera-based acquisition environment, along with the word and character level ground truth. The dataset can be used for training as well as testing of character recognition systems for camera-captured documents. Various benchmark tests are performed to analyze the behavior of different open source OCR systems on camera captured document images. Evaluation results show that the existing OCRs, which already get very high accuracies on scanned documents, fail on camera captured document images.
Using the presented camera-captured dataset, a novel character recognition system is developed which is based on a variant of recurrent neural networks, i.e., Long Short Term Memory (LSTM) that outperforms all of the existing OCR engines on camera captured document images with an accuracy of more than 95%.
Finally, this thesis provides details on various tasks that have been performed in the area closely related to information segmentation. This includes automatic analysis and sketch based retrieval of architectural floor plan images, a novel scheme for online signature verification, and a part-based approach for signature verification. With these contributions, it has been shown that part-based methods can be successfully applied to document image analysis.
Data is the new gold and serves as a key to answer the five W’s (Who, What, Where, When, Why) and How’s of any business. Companies are now mining data more than ever and one of the most important aspects while analyzing this data is to detect anomalous patterns to identify critical patterns and points. To tackle the vital aspects of timeseries analysis, this thesis presents a novel hybrid framework that stands on three pillars: Anomaly Detection, Uncertainty Estimation,
and Interpretability and Explainability.
The first pillar is comprised of contributions in the area of time-series anomaly detection. Deep Anomaly Detection for Time-series (DeepAnT), a novel deep learning-based anomaly detection method, lies at the foundation of the proposed hybrid framework and addresses the inadequacy of traditional anomaly detection methods. To the best of the author’s knowledge, Convolutional Neural Network (CNN) was used for the first time in Deep Anomaly Detection for Time-series (DeepAnT) to robustly detect multiple types of anomalies in the tricky
and continuously changing time-series data. To further improve the anomaly detection performance, a fusion-based method, Fusion of
Statistical and Deep Learning for Anomaly Detection (FuseAD) is proposed. This method aims to combine the strengths of existing wellfounded
statistical methods and powerful data-driven methods.
In the second pillar of this framework, a hybrid approach that combines the high accuracy of the deterministic models with the posterior distribution approximation of Bayesian neural networks is proposed.
In the third pillar of the proposed framework, mechanisms to enable both HOW and WHY parts are presented.
Embedded systems have become ubiquitous in everyday life, and especially in the automotive industry. New applications challenge their design by introducing a new class of problems that are based on a detailed analysis of the environmental situation. Situation analysis systems rely on models and algorithms of the domain of computational geometry. The basic model is usually an Euclidean plane, which contains polygons to represent the objects of the environment. Usual implementations of computational geometry algorithms cannot be directly used for safety-critical systems. First, a strict analysis of their correctness is indispensable and second, nonfunctional requirements with respect to the limited resources must be considered. This thesis proposes a layered approach to a polygon-processing system. On top of rational numbers, a geometry kernel is formalised at first. Subsequently, geometric primitives form a second layer of abstraction that is used for plane sweep and polygon algorithms. These layers do not only divide the whole system into manageable parts but make it possible to model problems and reason about them at the appropriate level of abstraction. This structure is used for the verification as well as the implementation of the developed polygon-processing library.
Nowadays, accounting, charging and billing users' network resource consumption are commonly used for the purpose of facilitating reasonable network usage, controlling congestion, allocating cost, gaining revenue, etc. In traditional IP traffic accounting systems, IP addresses are used to identify the corresponding consumers of the network resources. However, there are some situations in which IP addresses cannot be used to identify users uniquely, for example, in multi-user systems. In these cases, network resource consumption can only be ascribed to the owners of these hosts instead of corresponding real users who have consumed the network resources. Therefore, accurate accountability in these systems is practically impossible. This is a flaw of the traditional IP address based IP traffic accounting technique. This dissertation proposes a user based IP traffic accounting model which can facilitate collecting network resource usage information on the basis of users. With user based IP traffic accounting, IP traffic can be distinguished not only by IP addresses but also by users. In this dissertation, three different schemes, which can achieve the user based IP traffic accounting mechanism, are discussed in detail. The inband scheme utilizes the IP header to convey the user information of the corresponding IP packet. The Accounting Agent residing in the measured host intercepts IP packets passing through it. Then it identifies the users of these IP packets and inserts user information into the IP packets. With this mechanism, a meter located in a key position of the network can intercept the IP packets tagged with user information, extract not only statistic information, but also IP addresses and user information from the IP packets to generate accounting records with user information. The out-of-band scheme is a contrast scheme to the in-band scheme. It also uses an Accounting Agent to intercept IP packets and identify the users of IP traffic. However, the user information is transferred through a separated channel, which is different from the corresponding IP packets' transmission. The Multi-IP scheme provides a different solution for identifying users of IP traffic. It assigns each user in a measured host a unique IP address. Through that, an IP address can be used to identify a user uniquely without ambiguity. This way, traditional IP address based accounting techniques can be applied to achieve the goal of user based IP traffic accounting. In this dissertation, a user based IP traffic accounting prototype system developed according to the out-of-band scheme is also introduced. The application of user based IP traffic accounting model in the distributed computing environment is also discussed.
Computer-based simulation and visualization of acoustics of a virtual scene can aid during the design process of concert halls, lecture rooms, theaters, or living rooms. Because, not only the visual aspect of the room is important, but also its acoustics. In factory floors noise reduction is important since noise is hazardous to health. Despite the obvious dissimilarity between our aural and visual senses, many techniques required for the visualization of photo-realistic images and for the auralization of acoustic environments are quite similar. Both applications can be served by geometric methods such as particle- and ray tracing if we neglect a number of less important effects. By means of the simulation of room acoustics we want to predict the acoustic properties of a virtual model. For auralization, a pulse response filter needs to be assembled for each pair of source and listener positions. The convolution of this filter with an anechoic source signal provides the signal received at the listener position. Hence, the pulse response filter must contain all reverberations (echos) of a unit pulse, including their frequency decompositions due to absorption at different surface materials. For the room acoustic simulation a method named phonon tracing, since it is based on particles, is developed. The approach computes the energy or pressure decomposition for each particle (phonon) sent out from a sound source and uses this in a second pass (phonon collection) to construct the response filters for different listeners. This step can be performed in different precision levels. During the tracing step particle paths and additional information are stored in a so called phonon map. Using this map several sound visualization approaches were developed. From the visualization, the effect of different materials on the spectral energy / pressure distribution can be observed. The first few reflections already show whether certain frequency bands are rapidly absorbed. The absorbing materials can be identified and replaced in the virtual model, improving the overall acoustic quality of the simulated room. Furthermore an insight into the pressure / energy received at the listener position is possible. The phonon tracing algorithm as well as several sound visualization approaches are integrated into a common system utilizing Virtual Reality technologies in order to facilitate the immersion into the virtual scene. The system is a prototype developed within a project at the University of Kaiserslautern and is still a subject of further improvements. It consists of a stereoscopic back-projection system for visual rendering as well as professional audio equipment for auralization purposes.
Three dimensional (3d) point data is used in industry for measurement and reverse engineering. Precise point data is usually acquired with triangulating laser scanners or high precision structured light scanners. Lower precision point data is acquired by real-time structured light devices or by stereo matching with multiple cameras. The basic principle of all these methods is the so-called triangulation of 3d coordinates from two dimensional (2d) camera images.
This dissertation contributes a method for multi-camera stereo matching that uses a system of four synchronized cameras. A GPU based stereo matching method is presented to achieve a high quality reconstruction at interactive frame rates. Good depth resolution is achieved by allowing large disparities between the images. A multi level approach on the GPU allows a fast processing of these large disparities. In reverse engineering, hand-held laser scanners are used for the scanning of complex shaped objects. The operator of the scanner can scan complex regions slower, multiple times, or from multiple angles to achieve a higher point density. Traditionally, computer aided design (CAD) geometry is reconstructed in a separate step after the scanning. Errors or missing parts in the scan prevent a successful reconstruction. The contribution of this dissertation is an on-line algorithm that allows the reconstruction during the scanning of an object. Scanned points are added to the reconstruction and improve it on-line. The operator can detect the areas in the scan where the reconstruction needs additional data.
First, the point data is thinned out using an octree based data structure. Local normals and principal curvatures are estimated for the reduced set of points. These local geometric values are used for segmentation using a region growing approach. Implicit quadrics are fitted to these segments. The canonical form of the quadrics provides the parameters of basic geometric primitives.
An improved approach uses so called accumulated means of local geometric properties to perform segmentation and primitive reconstruction in a single step. Local geometric values can be added and removed on-line to these means to get a stable estimate over a complete segment. By estimating the shape of the segment it is decided which local areas are added to a segment. An accumulated score estimates the probability for a segment to belong to a certain type of geometric primitive. A boundary around the segment is reconstructed using a growing algorithm that ensures that the boundary is closed and avoids self intersections.
Adaptive Extraction and Representation of Geometric Structures from Unorganized 3D Point Sets
(2009)
The primary emphasis of this thesis concerns the extraction and representation of intrinsic properties of three-dimensional (3D) unorganized point clouds. The points establishing a point cloud as it mainly emerges from LiDaR (Light Detection and Ranging) scan devices or by reconstruction from two-dimensional (2D) image series represent discrete samples of real world objects. Depending on the type of scenery the data is generated from the resulting point cloud may exhibit a variety of different structures. Especially, in the case of environmental LiDaR scans the complexity of the corresponding point clouds is relatively high. Hence, finding new techniques allowing the efficient extraction and representation of the underlying structural entities becomes an important research issue of recent interest. This thesis introduces new methods regarding the extraction and visualization of structural features like surfaces and curves (e.g. ridge-lines, creases) from 3D (environmental) point clouds. One main part concerns the extraction of curve-like features from environmental point data sets. It provides a new method supporting a stable feature extraction by incorporating a probability-based point classification scheme that characterizes individual points regarding their affiliation to surface-, curve- and volume-like structures. Another part is concerned with the surface reconstruction from (environmental) point clouds exhibiting objects that are more or less complex. A new method providing multi-resolutional surface representations from regular point clouds is discussed. Following the applied principles of this approach a volumetric surface reconstruction method based on the proposed classification scheme is introduced. It allows the reconstruction of surfaces from highly unstructured and noisy point data sets. Furthermore, contributions in the field of reconstructing 3D point clouds from 2D image series are provided. In addition, a discussion concerning the most important properties of (environmental) point clouds with respect to feature extraction is presented.