Diese Dissertation befasst sich mit der sensorgeführten Regelung von Industrierobotern zur Konturverfolgung. Beispiele dafür sind das robotergestützte Nähen, Entgraten oder das Auftragen von Dichtmasse entlang von Schweißnähten. Beim Nähen und Entgraten müssen während der Verfolgung der Kontur bestimmte Kontaktkräfte an möglicherweise nachgiebigen Werkstücken eingehalten werden. Dabei ist es in modernen Fertigungsprozessen wichtig, die Bewegung des Roboters mit wenig Einrichtaufwand vorzugeben. Dazu werden Sensorsysteme eingesetzt, die Bildinformationen und Kraftmessungen verarbeiten, um den Roboter mit gewünschter Kontaktkraft entlang sichtbarer Konturen eines Werkstückes zu führen. Der Fokus dieser Arbeit ist die Fusion der Sensordaten, um die Vorteile der einzelnen Sensoren in einer Aufgabe zu vereinen. Es werden Messwerte eines Kraft-Momenten Sensors, einer Kamera, eines Beschleunigungssensors und der kartesischen Position und Orientierung des Roboters verwendet. Zuerst wird die Berechnung der kartesischen Roboterposition untersucht. Es wird ein Beobachter vorgestellt, um unter Verwendung eines Beschleunigungssensors die Präzision des Positionssignales zu erhöhen. Anschließend wird das Kamerasystem untersucht und ein Verfahren vorgestellt, um Geschwindigkeit, Position und Orientierung des robotergeführten Werkzeuges entlang der Kontur vorzugeben. Danach wird auf die Ermittlung von Kontaktkräften eingegangen, wobei die Kompensation von Trägheitskräften mittels Beschleunigungssensoren untersucht wird. Der letzte Abschnitt befasst sich mit der Verbindung von bildgestützter Konturverfolgung und Kraftregelung an nachgiebigen Werkstücken. Durch die Nachgiebigkeit des Werkstückmaterials verformt sich die Kontur bei Kontakt. Durch bildgestützte Konturverfolgung wird eine Anpassung an diese Verformung vorgenommen -- somit besteht über die Verformbarkeit des Werkstückes eine Kopplung zwischen den beiden Regelkreisen. Diese Kopplung wird gelöst, indem auf Basis eines dynamischen Modells der Umgebung eine Kompensation der Werkstückverformung berechnet wird. Die Modellparameter zur Kompensation werden durch online Identifikation ermittelt. / This thesis focuses on the sensor-guided control of industrial robots for contour-following. Examples include the robot-guided sewing, grinding or the application of sealant along weld seams. Grinding and sewing require a certain contact force while following a countour of a workpiece, whereas the worpieces might be compliant. Modern production processes require a fast and simple way to set up the motion of the robot for the required task. Therefore sensor systems are used, which process visual and tactile information to guide the tool at a desired contact force along visible contours of a workpiece. The focus of this work is the fusion of sensor data, used to benefit from the advantages of each of the individual sensors in one control scheme. I combine the measurements of a force-torque sensor, a camera, an acceleration sensor and of the Cartesian position of the robot. First, I introduce details on the calculation of the Cartesian robot position. I present an observer-based structure that uses an acceleration sensor to improve the precision of the robot position signal. Then, I analyze the camera system and present a control structure that adapts the position, orientation and velocity of the robot-guided tool along the contour. Thereafter, I show details of force measurement, whereas I compensate for inertial forces using an acceleration sensor. The last chapter addresses the combination of visual contour-following and force control on compliant workpieces. Under a certain contact force, the workpiece deforms due to its compliance. The position and orientation then is adapted to this deformed contour by visual control -- hence, there is a coupling between force and visual control. This coupling is solved by compensating for workpiece deformation using a dynamic model of the environment. The environmental parameters for compensation are identified online.

Cette thèse décrit quelques développements dans les techniques de calcul des amplitudes de diffusion en théorie supersymétrique de champ de jauge . L'accent est mis sur les relations de récurrence on-shell et sur l'utilisation de méthodes d'unitarité pour des calculs de boucle. En particulier, la récurrence on-shell est liée aux règles BCFW pour calculer les amplitudes de jauge au niveau des arbres. Les combinaisons de techniques de coupe d'unitarité et la récurrence sont utilisées pour calculer les amplitudes de boucle, et finalement, à partir des amplitudes, pour obtenir la fonction de corrélation énergie-énergie en théorie de super-Yang-Mills N = 4 à l'aide de la représentation de Mellin-Barnes. Dans le dernier chapitre, nous tentons de trouver un contour convergent pour les intégrales de Mellin Barnes en multi-dimension obtenu par une certaine approximation d'un contour de phase stationnaire.

This thesis examines an automated method to detect surface deformations on automotive panels for the purpose of quality control along a manufacturing assembly line. Automation in the automotive manufacturing industry is becoming more prominent, but quality control is still largely performed by human workers. Quality control is important in the context of automotive body panels as deformations can occur along the assembly line such as inadequate handling of parts or tools around a vehicle during assembly, rack storage, and shipping from subcontractors. These defects are currently identified and marked, before panels are either rectified or discarded. This work attempts to develop an automated system to detect deformations to alleviate the dependence on human workers in quality control and improve performance by increasing speed and accuracy. Some techniques make use of an ideal CAD model behaving as a master work, and panels scanned on the assembly line are compared to this model to determine the location of deformations. This thesis presents a solution for detecting deformations of various scales without a master work. It also focuses on automated analysis requiring minimal intuitive operator-set parameters and provides the ability to classify the deformations as dings, which are deformations that protrude from the surface, or dents, which are depressions into the surface. A complete automated deformation detection system is proposed, comprised of a feature extraction module, segmentation module, and classification module, which outputs the locations of deformations when provided with the 3D mesh of an automotive panel. Two feature extraction techniques are proposed. The first is a general feature extraction technique for 3D meshes using octrees for multi-resolution analysis and evaluates the amount of surface variation to locate deformations. The second is specifically designed for the purpose of deformation detection, and analyzes multi-resolution cross-sections of a 3D mesh to locate deformations based on their estimated size. The performance of the proposed automated deformation detection system, and all of its sub-modules, is tested on a set of meshes which represent differing characteristics of deformations in surface panels, including deformations of different scales. Noisy, low resolution meshes are captured from a 3D acquisition, while artificial meshes are generated to simulate ideal acquisition conditions. The proposed system shows accurate results in both ideal situations as well as non-ideal situations under the condition of noise and complex surface curvature by extracting only the deformations of interest and accurately classifying them as dings or dents.

This thesis examines an automated method to detect surface deformations on automotive panels for the purpose of quality control along a manufacturing assembly line. Automation in the automotive manufacturing industry is becoming more prominent, but quality control is still largely performed by human workers. Quality control is important in the context of automotive body panels as deformations can occur along the assembly line such as inadequate handling of parts or tools around a vehicle during assembly, rack storage, and shipping from subcontractors. These defects are currently identified and marked, before panels are either rectified or discarded. This work attempts to develop an automated system to detect deformations to alleviate the dependence on human workers in quality control and improve performance by increasing speed and accuracy. Some techniques make use of an ideal CAD model behaving as a master work, and panels scanned on the assembly line are compared to this model to determine the location of deformations. This thesis presents a solution for detecting deformations of various scales without a master work. It also focuses on automated analysis requiring minimal intuitive operator-set parameters and provides the ability to classify the deformations as dings, which are deformations that protrude from the surface, or dents, which are depressions into the surface. A complete automated deformation detection system is proposed, comprised of a feature extraction module, segmentation module, and classification module, which outputs the locations of deformations when provided with the 3D mesh of an automotive panel. Two feature extraction techniques are proposed. The first is a general feature extraction technique for 3D meshes using octrees for multi-resolution analysis and evaluates the amount of surface variation to locate deformations. The second is specifically designed for the purpose of deformation detection, and analyzes multi-resolution cross-sections of a 3D mesh to locate deformations based on their estimated size. The performance of the proposed automated deformation detection system, and all of its sub-modules, is tested on a set of meshes which represent differing characteristics of deformations in surface panels, including deformations of different scales. Noisy, low resolution meshes are captured from a 3D acquisition, while artificial meshes are generated to simulate ideal acquisition conditions. The proposed system shows accurate results in both ideal situations as well as non-ideal situations under the condition of noise and complex surface curvature by extracting only the deformations of interest and accurately classifying them as dings or dents.

The expected rise in aviation demand requires the reduction of the environmental impacts that impede this desired growth, such as fuel burn, emissions, and airport noise. A number of current technology programs attempt to identify, evaluate, and select the environmental technology solutions for the coming decades. Fleet-level evaluation will be essential to deciding between various technology options because it provides a system-level assessment that clarifies the effect of operational and policy variables. Fleet-level modeling in general, introduces various complexities, and detailed fleet-level models require significant time and computing resources to execute. With a large number of potential technology options available for assessment, a full detailed analysis of the technology space is infeasible. Therefore, a simplified fleet-level environmental evaluation methodology is required to select scenarios to carry forward for detailed modeling. Capabilities such as the Global and Regional Environmental Aviation Tradeoff (GREAT) tool, have achieved rapid simplified fleet-level analysis for fuel burn and emissions, but currently lack a satisfactory generic framework to evaluate fleet-level noise. The primary objective of this research is to formulate and implement a generic fleet-level noise methodology that allows decision makers to analyze the fleet-level impact of many technology scenarios on the quantity of noise, and also its distribution about certain airport types. This information can be leveraged to provide screening assessments of technology impacts earlier in the decision-making process, reserving more sophisticated modeling techniques for the most promising scenarios. The capability gaps identified are addressed by the development of a rapid generic fleet-level noise model that captures basic airport noise contour shape and contour area, a categorization of airports with respect to their operational and infrastructure characteristics, and the development of shape metrics that enable rapid classification and comparison of contour shapes. Once the capability gaps were addressed, the resultant System-Wide Assessment of Noise (SWAN) methodology was implemented via use cases to demonstrate the application of the methodology, examining the introduction of a set of possible near-term (N+1) future technologies into the forecast. While these examples are simplified and notional, they demonstrate the types of analyses and investigations that can be performed with the SWAN methodology, providing answers regarding the impact of technologies on contour shapes. The development, verification, validation, and demonstration of these capabilities complete a framework for evaluating fleet-level noise at the screening-level that retains the ability to capture and effectively discuss shape information beyond the capability of current screening-level noise evaluation techniques. By developing a rapid generic fleet-level noise model, a set of Generic Airports, and metrics that objectively quantify and describe shape, decision-makers can access greater levels of information, including the critical facet of contour shape in fleet-level airport noise.

Prostate segmentation is a required step in determining the volume of a prostate, which is very important in the diagnosis and the treatment of prostate cancer. In the past, radiologists manually segment the two-dimensional cross-sectional ultrasound images. Typically, it is necessary for them to outline at least a hundred of cross-sectional images in order to get an accurate estimate of the prostate's volume. This approach is very time-consuming. To be more efficient in accomplishing this task, an automated procedure has to be developed. However, because of the quality of the ultrasound image, it is very difficult to develop a computerized method for defining boundary of an object in an ultrasound image. The goal of this thesis is to find an automated segmentation algorithm for detecting the boundary of the prostate in ultrasound images. As the first step in this endeavour, a semi-automatic segmentation method is designed. This method is only semi-automatic because it requires the user to enter four initialization points, which are the data required in defining the initial contour. The discrete dynamic contour (DDC) algorithm is then used to automatically update the contour. The DDC model is made up of a set of connected vertices. When provided with an energy field that describes the features of the ultrasound image, the model automatically adjusts the vertices of the contour to attain a maximum energy. In the proposed algorithm, Mallat's dyadic wavelet transform is used to determine the energy field. Using the dyadic wavelet transform, approximate coefficients and detailed coefficients at different scales can be generated. In particular, the two sets of detailed coefficients represent the gradient of the smoothed ultrasound image. Since the gradient modulus is high at the locations where edge features appear, it is assigned to be the energy field used to drive the DDC model. The ultimate goal of this work is to develop a fully-automatic segmentation algorithm. Since only the initialization stage requires human supervision in the proposed semi-automatic initialization algorithm, the task of developing a fully-automatic segmentation algorithm is reduced to designing a fully-automatic initialization process. Such a process is introduced in this thesis. In this work, the contours defined by the semi-automatic and the fully-automatic segmentation algorithm are compared with the boundary outlined by an expert observer. Tested using 8 sample images, the mean absolute difference between the semi-automatically defined and the manually outlined boundary is less than 2. 5 pixels, and that between the fully-automatically defined and the manually outlined boundary is less than 4 pixels. Automated segmentation tools that achieve this level of accuracy would be very useful in assisting radiologists to accomplish the task of segmenting prostate boundary much more efficiently.

This thesis presents the analysis of artistic techniques of paintings from the Precisionist movement and the implementation of the results of the analysis in the creation of three new works of art using digital media. Artists working in digital media express features of pre-digital artistic movements with varying degrees of adherence to principles, intentions, and awareness. This thesis seeks to create a bridge between the recognition of common features of Precisionist works and the expression of those elements in new works through the use of a system of analysis, interpretation, and translation. One outcome of this thesis is the description of a methodology for interpretation and translation that can be applied to other art movements. The Precisionist period within the Modern Art movement has both a historical importance in the world of art and a thematic relevance to popular uses of digital media ? specifically the representation of meaning and mood derived from industrial settings. Its influences can be traced from cubist, futurist, and constructivist art, as well as influencing the development of surrealism. It is considered the first solely American movement within Modern Art. Charles Sheeler's work plays a key role in the visual analysis portion of this research. Sheeler's work offers examples for applying 2D precisionist artistic style as aesthetic inspiration in creating a three-part production of 3D digital and video work. Work from precisionist artists Charles Demuth and Edmund Lewandowski also contribute some unique artistic characteristics considered during the analytical portion of this study. The new artistic works proposed include: (1) a linear, live-action short video with post-production manipulation; (2) a linear, 3D animated work; and (3) a non-linear, interactive 3D game environment.

No description available.

This thesis tackles the visual tracking problem as a target contour estimation problem in the face of corrupted measurements. The major aim is to design robust recursive curve filters for accurate contour-based tracking. The state-space representation adopted comprises of a group component and a shape component describing the rigid motion and the non-rigid shape deformation respectively; filtering strategies on each component are then decoupled. The thesis considers two implicit curve descriptors, a classification probability field and the traditional signed distance function, and aims to develop an optimal probabilistic contour observer and locally optimal curve filters. For the former, introducing a novel probabilistic shape description simplifies the filtering problem on the infinite-dimensional space of closed curves to a series of point-wise filtering tasks. The definition and justification of a novel update model suited to the shape space, the derivation of the filtering equations and the relation to Kalman filtering are studied. In addition to the temporal consistency provided by the filtering, extensions involving distributed filtering methods are considered in order to maintain spatial consistency. For the latter, locally optimal closed curve filtering strategies involving curve velocities are explored. The introduction of a local, linear description for planar curve variation and curve uncertainty enables the derivation of a mechanism for estimating the optimal gain associated to the curve filtering process, given quantitative uncertainty levels. Experiments on synthetic and real sequences of images validate the filtering designs.

This thesis deals with image segmentation problems that arise in various computer vision related fields such as medical imaging, satellite imaging, video surveillance, recognition and robotic vision. More specifically, this thesis deals with a special class of image segmentation technique called Snakes or Active Contour Models. In active contour models, image segmentation is posed as an energy minimization problem, where an objective energy function (based on certain image related features) is defined on the segmenting curve (contour). Typically, a gradient descent energy minimization approach is used to drive the initial contour towards a minimum for the defined energy. The drawback associated with this approach is that the contour has a tendency to get stuck at undesired local minima caused by subtle and undesired image features/edges. Thus, active contour based curve evolution approaches are very sensitive to initialization and noise. The central theme of this thesis is to develop techniques that can make active contour models robust against certain classes of local minima by incorporating global information in energy minimization. These techniques lead to energy minimization with global considerations; we call these models -- 'Non-local active contours'. In this thesis, we consider three widely used active contour models: 1) Edge- and region-based segmentation model, 2) Prior shape knowledge based segmentation model, and 3) Motion segmentation model. We analyze the traditional techniques used for these models and establish the need for robust models that avoid local minima. We address the local minima problem for each model by adding global image considerations.