Persistics is an advanced framework for processing wide-area aerial surveillance video. This framework handles the tasks of data collection, stitching of multi-sensor imagery, image registration and stabilization, motion tracking, and compression. As the technology for image sensor sizes improves, significant improvements in compression techniques are necessary in order to make full use of the data. Because the information of interest in such video is naturally moving, point-like targets, the applicability of foveated coding to the compression problem is an interesting question. Foveated coding, a compression technique that was designed to be perceptually optimal for the human visual system, has several components that are appropriate to the persistics compression problem. Foveation is applied in several different scenarios and methods to persistics data. As foveation can make good use of the persistics tracker data, a problem affecting tracker performance is explored as well. The multi-sensor stitching component of persistics can generate artifacts that reduce the effectiveness of the tracker. A method for characterizing, detecting, and correcting such artifacts is desirable. These three concepts are explored, and a method for detection is developed. Components of these algorithms were absorbed into a more general framework for artifact correction. / text

In this thesis, the applications of image stitching and object insertion are considered and two gradient based approaches offering solutions are proposed. An essential part of the proposed methods is obtaining an image from a given gradient data set. This is done using an existing Haar wavelet based reconstruction technique, which consists of two main steps. First, the Haar wavelet decomposition of the image to be reconstructed is obtained directly from a given gradient. Second, the image is obtained using Haar wavelet synthesis. In both stitching and object insertion applications considered, the gradient from which the image must be reconstructed is a non-conservative vector field and this requires adding an iterative Poisson solver at each resolution level, during the synthesis step of the reconstruction technique. The performance of the reconstruction algorithm is evaluated by comparing it with other existing techniques, in terms of solution accuracy and computation speed. The proposed image stitching technique consists of three main parts: registering the images to be combined, blending their gradients over a region of interest and obtaining a composite image from a gradient. The object insertion technique considers the images registered and has two main stages: gradient blending of images in a region of interest and recovering an image from the gradient. The performance of the stitching algorithm is evaluated visually, by presenting the results produced to combine images with varying orientation, scales, illumination, and color conditions. Experimental results illustrate both the stitching and the insertion techniques proposed, and indicate that they yield seamless composite images. / Graduate

Integrated camera systems for increasing safety and maneuverability are becoming increasingly common for heavy vehicles. One problem with heavy vehicles today is that there are blind spots where the driver has no or very little view. There is a great demand on increasing the safety and helping the driver to get a better view of his surroundings. This can be achieved by a sophisticated camera system, using cameras with wide field of view, that could cover dangerous blind spots. This master thesis aims to investigate and develop a prototype solution for a camera system consisting of two fisheye cameras. The solution covers both hardware choices and software development including camera calibration and image stitching. Two different fisheye camera calibration toolboxes are compared and their results discussed, with the aim to find the most suitable for this application. The result from the two toolboxes differ in performance, and the result from only one of the toolboxes is sufficient for image stitching.

With increased mobile hardware capabilities, improved UAVs and modern algorithms, accurate maps can be created in real-time by capturing overlapping photographs of the ground. A method for mapping that can be used is to position photos by relying purely on the GPS position and altitude. However, GPS inaccuracies will be visible in the created map. In this paper, we will instead present a method for aligning the photos correctly with the help of feature matching. Feature matching is a well-known method which analyses two photos to find similar parts. If an overlap exists, feature matching can be used to find and localise those parts, which can be used for positioning one image over the other at the overlap. When repeating the process, a whole map can be created. For this purpose, we have also evaluated a selection of feature detection and matching algorithms. The algorithm found to be the best was SIFT with FLANN, which was then used in a prototype for creating a complete map of a forest. Feature matching is in many cases superior to GPS positioning, although it cannot be fully depended on as failed or incorrect matching is a common occurrence.

<div>Many Engineering structures, including electronic component assemblies, are inherently multi-scale in nature. These structures often experience complex local nonlinear behavior such as plasticity, damage or fracture. These local behaviors eventually lead to the failure at the macro length scale. Connecting the behavior across the length scales to develop an understanding of the failure mechanism is important for developing reliable products.</div><div><br></div><div>To solve multi-scale problems in which the critical region is much smaller than the entire structure, an iterative solution approach based on domain decomposition techniques is proposed. Two independent models are constructed to model the global and local substructures respectively. The unbalanced force at the interface is iteratively reduced to ensure force equilibrium of the overall structure in the final solution. The approach is non-intrusive since only nodal values on the interface are transferred between the global and local models. Solution acceleration using SR1 and BFGS updates is also demonstrated. Equally importantly, the two updates are applied in a non-intrusive manner, meaning that the technique is implemented without needing access to the codes using which the sub-domains are analyzed. Code- and mesh-agnostic solutions for problems with local nonlinear material behavior or local crack growth are demonstrated. Analysis in which the global and local models are solved using two different commercial codes is also demonstrated.</div><div><br></div><div>Engineering analysis using numerical models are helpful in providing insight into the connection between the structure, loading history, behavior and failure. Specifically, Isogeometric analysis (IGA) is advantageous for engineering problems with evolving geometry compared to the traditional finite element method (FEM). IGA carries out analysis by building behavioral approximations isoparametrically on the geometrical model (commonly NURBS) and is thus a promising approach to integrating Computer-Aided Design (CAD) with Computer-Aided Engineering (CAE).</div><div><br></div><div>In enriched isogeometric Analysis (EIGA), the solution is enriched with known behavior on lower dimensional geometrical features such as crack tips or interfaces. In the present research, enriched field approximation techniques are developed for the application of boundary conditions, coupling patches with non-matching discretizations and for modeling singular stresses in the structure.</div><div><br></div><div>The first problem solution discussed is to apply Dirichlet and Neumann boundary conditions on boundary representation (B-rep) CAD models immersed in an underlying domain of regular grid points. The boundary conditions are applied on the degrees of freedom of the lower dimensional B-rep part directly. The solution approach for the immersed analysis uses signed algebraic level sets constructed from the B-rep surfaces to blend the enriched</div><div>field with the underlying field. The algebraic level sets provide a surrogate for distance, are non-iteratively (or algebraically) computed and allow implicit Boolean compositions.</div><div><br></div><div>The methodology is also applied to couple solution approximations of decomposed patches by smoothly blending incompatible geometries to an arbitrary degree of smoothness. A parametrically described frame or interface is introduced to “stitch” the adjacent patches. A hierarchical blending procedure is then developed to stitch multiple unstructured patches including those with T-junctions or extraordinary vertices.</div><div><br></div><div>Finally, using the EIGA technique, a computational method for analyzing general multimaterial sharp corners that enables accurate estimations of the generalized stress intensity factors is proposed. Explicitly modeled geometries of material junctions, crack tips and deboned interfaces are isogeometrically and hierarchically enriched to construct approximations with the known local behavior. specifically, a vertex enrichment is used to approximate the asymptotic field near the re-entrant corner or crack tip, Heaviside function is used to approximate the discontinuous crack face and the parametric smooth stitching technique is used to approximate the behavior across material interface. The developed method allows direct extraction of generalized stress intensity factors without needing a posteriori evaluation of path independent integrals for decisions on crack propagation. The numerical implementation is validated through analysis of a bi-material corner, interface crack and growth of an inclined crack in a homogeneous solid. The developed procedure demonstrates rapid convergence to the solution stress intensity factors with relatively fewer degrees of freedom, even with uniformly coarse discretizations.</div>

In this thesis, we present three novel algorithms. The first algorithm is a method of identifying numerical landmarks (a definition coined in this thesis). The second algorithm uses the projection of image regions onto x- and y- axes and the matching of the resulting 1D projections to determine an overall 2D translation for use in registration. The third algorithm aligns SEM images of successive layers of a semiconductor device by first extracting the positions of vias in the lower layer, and then searching for the best translation for subsets of vias such that they all or mostly connect to metalization in the upper layer. / To acquire high-resolution Scanning Electron Microscopy (SEM) images over wide areas, we must acquire several images ``tiling'' the surface and assemble them into a single composite image, using a process called image stitching. While for some applications, stitching is now routine, SEM mosaics of semiconductors pose several challenges: (1) by design, the image features (wire, via and dielectric) are highly repetitive, (2) the overlap between image tiles is small, (3) sample charging causes intensity variation between captures of the same region, and (4) machine instability causes non-linear deformation within tiles and between tiles. In this study, we compare the accuracy and computational cost of three well-known pixel-based techniques: Fast Fourier Transform (FFT), Sum of Squared Differences (SSD), and Normalized Cross Correlation (NCC). We compare well-known 2D algorithms, as well as novel projection-onto-1D versions. The latter reduces the computational complexity from O(n^2) to O(n), where n is the number of pixels, without loss of accuracy, and in some cases, with greater accuracy. Another approach to reducing the computational complexity of image alignment is to compare isolated landmarks, rather than pixels. In semiconductor images, there are no natural fiducials and adding them would destroy the information required to reconstruct their circuits, so we introduce a new class of landmarks which we call numerical landmarks. Related to Harris corners, the novel numerical landmarks are insensitive to brightness variations and noise. Finally, we consider the alignment problem between layers of image mosaics. Unlike in the ``horizontal'' directions, the vertical dimension is only sparsely sampled. Consequently, image features and landmarks cannot be used for alignment. Instead, we must rely on the relationship between vias (through-plane metalization) and wires (in-plane metalization), and we have developed a novel algorithm for matching vias in the lower layer with wires above, and use this to align subimages. / Thesis / Doctor of Philosophy (PhD) / Applications in materials science often require the acquisition of images of semiconductor computer chips at very high resolution. Using cameras with even tens of millions of pixels might not give us enough resolution over a wide field of view. One approach is to acquire several images of parts of the sample at high magnification and assemble them into a single composite image. This way, we can preserve the high resolution over a wide area. Algorithms developed for assembling the composite image are known as tiling or mosaicing. This whole process is known as image stitching (and includes image registration). In this thesis, we develop specialized algorithms suited for the 2D stitching of semiconductor images, including the generalization to 3D. This case is challenging because slight alignment errors may completely change the reconstructed circuit, and the images contain both repeated patterns (such as many parallel wires) and changes in brightness and distortions caused by the scanning device.

No description available.

Event Detection is a process of identifying terrain flatness from which localized events such as potholes in the terrain surface can be found and is an important tool in pavement health monitoring and vehicle performance inspection. Repeated detection of terrain surfaces over an extended period of time can be used by highway engineers for long term road health monitoring. An accurate terrain map can allow maintenance personnel for identifying deterioration in road surface for immediate correction. Additionally, knowledge of the events in terrain surface can be used to predict the performance the vehicles would experience while traveling over it. Event detection is composed of two processes: event edging and stitching edges to events. Edge detection is a process of identifying significant localized changes in the terrain surface. Many edge detection methods have been designed capable of capturing edges in terrain surfaces. Gradient searches are frequently used in image processing to recover useful information from images. The issue with using a gradient search method is that it returns deterministic values resulting in edges which are less precise. In order to predict the precision of the terrain surface, the individual nodal probability densities must be quantified and finally combined for the precision of terrain surface. A Comparative Nodal Uncertainty Method is developed in this work to detect edges based on the probability distribution of the nodal heights within some local neighborhood. Edge stitching is developed to group edges to events in a correct sequence from which an event can be determined finally. / Master of Science

This project explores the design and implementation of synchronized Pan-Tilt-Zoom (PTZ) camera control for panoramic imaging, with a specific focus on enhancing surveillance systems in air traffic management. The motivation for this study comes from the need for better monitoring in air traffic control, where panoramic views can greatly improve situational awareness and safety. The main challenge is coordinating multiple PTZ cameras to capture and stitch images, creating a comprehensive panoramic view despite individual camera limitations. The study uses simulation to test synchronization and image stitching techniques, ensuring camera alignment and seamless panoramic images. Results indicate the system's robustness and potential for real-world applications, though future validation with physical hardware is necessary.

<p>A new approach for object replacement in 3D space is presented. Introducing a technique that replaces the older two dimensional (2D) based facial replacement method performed by compositing artist in motion picture productions and video commercial industry.</p><p>This method uses 4 digital video cameras filming an actor from 360 degrees, the cameras are placed with 90 degrees in between, the video footage acquired is then used to produce a 3D video texture consisting of video segments taken from different angles representing the object from 3D point of view.</p><p>The video texture is then applied to a 3D modelled head matching the geometry of the original object.</p><p>Offering the freedom of showing the object from any point of view from 3D space, which is not possible using the current two dimensional method where the actormust at all time face the camera.</p><p>The method is described in details with images showing every stage of the process.</p><p>Results are presented as still frames taken from the final video footage and as a video file demonstrating them.</p>