Learning-based multi-view stereo (MVS) has shown promising results in the domain of general 3D reconstruction. However, no work before this thesis has applied learning-based MVS to urban 3D reconstruction from satellite images. In this thesis, learning-based MVS is used to infer depth maps from satellite images. Models are trained on both synthetic and real satellite images from Las Vegas with ground truth data from a high-resolution aerial-based 3D model. This thesis also evaluates different methods for reconstructing digital surface models (DSM) and compares them to existing satellite-based 3D models at Maxar Technologies. The DSMs are created by either post-processing point clouds obtained from predicted depth maps or by an end-to-end approach where the depth map for an orthographic satellite image is predicted.  This thesis concludes that learning-based MVS can be used to predict accurate depth maps. Models trained on synthetic data yielded relatively good results, but not nearly as good as for models trained on real satellite images. The trained models also generalize relatively well to cities not present in training. This thesis also concludes that the reconstructed DSMs achieve better quantitative results than the existing 3D model in Las Vegas and similar results for the test sets from other cities. Compared to ground truth, the best-performing method achieved an L1 and L2 error of 14 % and 29 % lower than Maxar's current 3D model, respectively. The method that uses a point cloud as an intermediate step achieves better quantitative results compared to the end-to-end system. Very promising qualitative results are achieved with the proposed methods, especially when utilizing an end-to-end approach.

This thesis describes a distributed system of cameras for visually tracking feature points in 3D. The concept of a network of cameras is introduced. Two or more calibrated cameras from a network of cameras are used to triangulate the location of a point in 3D based on camera positions and pan/tilt angles. A survey of methods for interior and exterior calibration is provided and a method suitable for multiple cameras in arbitrary positions has been implemented. A low-cost camera unit has been designed using off-the-shelf components that include a small CCD board camera and two servo-controlled mechanisms for pan and tilt. Experimental results demonstrate the performance of a network of cameras. / Thesis / Master of Engineering (ME)

During an inventory of Ör church’s roof structure in 2021, it was discovered that the roofstructure above the church’s nave contains a large amount of reused structural parts froman earlier roof structure. The reused structural parts are most likely originated from Örchurch’s early medieval roof structure.This thesis is an attempt to reconstruct the early medieval roof structure with the use of3D-laser scanning. The goal is to present a possible shape and 3D-model of the earlymedieval roof structure. The existing roof structure will also be documented. One aim ofthe thesis is to develop and evaluate the method of using 3D-laser scanning for thispurpose.The reconstruction resulted in eight different shapes and 3D-models of the early medievalroof structure. The documentation of the existing roof structure above the nave resulted inupdated section, plan and detailed drawings.The method of using 3D-laser scanning gave a detailed point cloud of the existing roofstructure. The point cloud had low density points in certain parts of the roof structurewhich is a consequence from the implementation of the 3D laser scanning. For thismethod a field study and planning in advance is recommended for best possible result.

Bakgrunden till arbetet ligger i de lagändringar som trädde i kraft den 1 maj 2009. Med hjälp av dessa blir det möjligt att inrätta ägarlägenheter som en ny boendeform i Sverige. Vi vill klargöra vad som gäller för att få inrätta ägarlägenheter i befintlig bebyggelse samt utreda hur erfarenheter, kunskaper och förväntningarna på den nya boendeformen ser ut.Vid ett flertal tillfällen sedan 1990-talet har frågan om att få inneha enskilda lägenheter i flerbostadshus med äganderätt s.k. ägarlägenheter varit aktuell. År 2004 infördes möjlighet till tredimensionell fastighetsindelning, vilket är avgörande för att kunna skapa ägarlägenheter. Boendeformen är vanlig utomlands. Lagändringarna för hur en ägarlägenhet i Sverige ska ses påminner mest om Danmarks ejerlejigheder. En ägarlägenhetsfastighet ska innehålla endast en bostadslägenhet, som vanligtvis utgörs av lägenhetsutrymmet inklusive ett ytskikt, och ingå i en sammanhållen enhet om minst tre sådana. De kan bildas antingen i nyuppförda byggnader eller genom ombildning av äldre byggnader där det aktuella utrymmet inte till någon del har använts som bostadslägenhet under de senaste åtta åren.Vid fastighetsbildning måste Lantmäteriet ta i beaktande att fastigheten som bildas blir varaktigt lämpad för sitt ändamål samt rättsligt försäkrad tillgång till vatten och avlopp. Det bästa sättet att försäkra en byggnad dessa tillgångar är genom bildandet av en gemensamhetsanläggning eller med fastighetsbildningsservitut. En gemensamhets-anläggning förvaltas av en samfällighetsförening där fastighetsägarna är andelsägare i föreningen.Utifrån intervjuerna kan man konstatera att erfarenheten och kunskapen om ämnet är begränsad. Förväntningarna på den nya boendeformen är högt ställda. Man tror bl.a. att tillkomsten kan leda till minskat tryck på bostadsmarknaden och ge möjlighet för fastighetsägare att utnyttja onyttiga ytor på ett bättre sätt. / The background to the work lies in the legislative changes which came into effect May 1, 2009. The changes will make it possible to establish ownership condominiums as a new form of housing in Sweden. We want to clarify how to adapt ownership condo-miniums in existing settlements, and examine how experience, knowledge and expectations of the new housing arrangements are. On several occasions since the 1990s, questions have arisen to establish ownership of an individual apartment inside an apartment-complex. In 2004, three dimensional real estate division was introduced, which is essential to create ownership condominiums. This housing form is used internationally. The legislative amendments on the procedures of condominium ownership in Sweden will be most reminiscent of Denmarks ejerlejigheder. A condominium property should contain only one dwelling, which usually consists of the apartment space and a surface layer, and it should be a part of a cohesive unit of at least three such. They can be established either in new buildings or by conversion of older buildings where no space has been used as a dwelling in the past eight years. During a property formation the Lantmäteriet must take into consideration that the property formed becomes durable. It should be suitable for its purpose and have insured access to water and sewage systems. The best way to assure that a building gains these assets is through the formation of a single-unit establishment or property assessments. Jointly-owned property areas are managed by a cooperative were the owners of the condominiums are members.From the interviews, it can be said that the experience and knowledge of the subject is limited. There are high expectations of the new housing form. It is believed among other things that the appearance may lead to less pressure on the current real-estate market. This will also create new opportunities for property-owners to use vacant space in a better way.

<p> One of the most difficult and interesting aspects of the physics of collapse and outflow formation, as well as the evolution of the protostellar disk, is the role of hydromagnetic forces. However, magnetic fields are only coupled to the charged species present in poorly ionized molecular clouds. Ambipolar diffusion-the process by which magnetic fields "slip" in poorly ionized gas-strongly affects the initial cloud as well as the final observable structure through collisional heating. Also, as the gas becomes opaque to cosmic rays, the ionized structure of the accreting gas may become more complex, leading to a neutral 'dead zone' in a layered accretion disk (vital in determining planet masses in planet formation theories (Matsumura & Pudritz, 2005)). We omit possible effects of ionizing radiation in these early stages of formation. </p> <p> In this thesis, we perform fully 3D simulations (using the FLASH AMR code) and have implemented ambipolar diffusion in the MHD module of the code in addition to a broad treatment of cooling (Banerjee et al., 2006). This has allowed us to track the ionized gas and magnetic fields properly from the beginning of collapse down to the onset of outflows. We find that high accretion rates persist on the order 1 of 10-3 M0 yr-(where the core mass has reached about 0.1 M0 ) due to efficient extraction of angular momentum through magnetic processes. Magnetic braking is reduced by about 3/4 in the initial collapse relative to an ideal collapse of same initial conditions. This, with a reduction in magnetic pressure in the disk, leads to an increased rate of fragmentation. One of the major new results of this work is the discovery that outflows from disks still occur even in the presence of ambipolar diffusion. Surprisingly, they are initiated even earlier than outflows from idealized, completely ionized disks. They are generated by a magnetic tower mechanism at central densities of 1012 cm-3, as effective ram pressure on the wound up toroidal field is reduced, allowing it to push away from the disk earlier. </p> <p> We have also shown that the formation of a dead zone in these early stages is dependent on shielding of cosmic rays, in the absence of which a decoupled zone in the disk midplane forms. This region, where the accreting gas is effectively decoupled from the magnetic field, extends 10 AU in radius and (2-3) AU in height from the midplane. The global magnetic field threading such a complex accretion disk shows a dragged out structure, as coupled surface layers of the disk pull in the field. The disk is puffy due to drift heating and the initial stages of the outflow pushing out into the ambient medium. However, overall magnetic field build-up is still efficient, as values of the magnetic field in the disk are only reduced by half. </p> / Thesis / Master of Science (MSc)

<p> A new area-based mesh simplification algorithm is described. The proposed algorithm removes the center vertex of a polygon which consists of n ≥ 3 faces and represents that polygon with n - 2 faces. A global search method is adapted that iteratively determines which vertex is to be removed using the proposed area-based distortion measurement. Although the global search method requires more computations compared to a local search method, it guarantees better quality of approximation. Various re-triangulations are also considered to improve the perceptual quality of the final approximation.</p> <p> From multiple re-triangulations, one with minimum distortion is selected to represent the original mesh. Experimental results demonstrate the performance of the proposed algorithm for data reduction while maintaining the quality of the rendered objects. The performance of multiple description decoder when not all descriptions are available depends on the decoding strategy. By approximating lost description the distortion can be reduced. When decoder reconstructs input source without having all descriptions different methods exist to approximate the lost description. We proposed two side decoding algorithms. The proposed side decoders are based on diagonal element and the probability of input source. When low bit-rate and complicated index assignment matrix are used the side decoder based on probability of input source is recommendable. To approximate the lost description we compare the performance of standard decoding method with the performance of proposed methods. A trade-off between the performance of decoder and computational complexity exists.</p> <p> An error concealment algorithm is proposed based on flow of facial expression to improve communication of animated facial data over a limited bandwidth channel with error. Facial expression flow is tracked using dominant muscles which are those with maximum change between two successive frames. By comparing the dominant muscle data with the predetermined expression information table, facial expression flow is determined. The receiver uses linear interpolation and the information on facial expression flow to interpolate the erroneous facial animation data. For objective comparison, a distortion measurement tool, which compares two 3D objects based on point-to-point difference, is introduced. Experimental results are provided to show that the proposed error concealment method improves the quality of an animated face communications.</p> / Thesis / Master of Applied Science (MASc)

No / Computational Fluid Dynamics (CFD) represents a useful tool to study natural currents in the rivers and estuaries with erosive materials; therefore, it is always in the keen interest for scientists to further study and advance it, especially when numerical model has the advantages compared to actual laboratory experiment in terms of cost, time, and restrictions on conditions of the physical models and field collections. The present study deals with the hydro-morphological investigation and numerical modeling of a group of vertically stationed parallel groynes using FLOW-3D commercial software. To validate the results of the FLOW-3D simulation, it has been compared to the experiments from literature. Besides, a SSIIM 2.0 software has also been employed to compare with some of the FLOW-3D results. It was found that the accuracy of the FLOW-3D model influenced by the approach Froude number and the critical velocity ratio (Uavg/Ucr). Even though it underestimated the measured scour depth (due to complex and intense vortices, which reduce the accuracy of the numerical models), but general results from the model have reproduced the measured data well.

Design review is a process of reviewing construction design documents to ensure that they reflect the owner's design intent, and are accurate in describing the owner's desired building or facility. Information generation becomes more intensive as the design stage progresses. The use of valuable information during design review stage can lead to a more comprehensive and high quality design, and a building or facility that is constructible, and within the intended budget. However, in current design practices, valuable design review information is scattered, ineffectively placed, and is not used efficiently. The design review process will be more efficient if this valuable information is integrated and centralized. The author developed a framework to improve the design review process by incorporating a centralized repository of design review information and 3D CAD model, in an interactive Virtual Environment (VE). To develop the framework, the author used Action Research style where he identified and confirmed the design review problem area, promoted the potential solutions to the problem, and developed a prototype. In gathering and analyzing the data for the research, the author used the synthesis of three methods. They include review of literature, a case study (interviews with industry personnel and content analysis of design review documents), and dissemination of the author's progressive findings in conferences, conference proceedings and journal publications. From his findings, the author developed the framework to improve the design review process by using information filtering based on context-aware concept, coupled with the benefits of a VE. The required design review information in the form of textual, numerical and geometric information is processed (queried, retrieved and stored). The author defined four contexts for information filtering: discipline-centric, task-centric, object-centric, and location-centric. IF-THEN rules are used to trigger the processing of the required design review information and present it to the design reviewer in a VE. A low cost 3D Game Engine is used as the enabling development tool to develop a work-in-progress (WIP) prototype design review application in a VE. / Ph. D.

If “a picture is worth a thousand words,” then a video may be worth a thousand pictures. Videos have been increasingly used in multiple applications, including surveillance, teleconferencing, learning and experience sharing. Since a video captures a scene from a particular viewpoint, it can often be understood better if presented within a larger spatial context. We call such interactive visualizations that combine videos with their spatial context "Contextualized Videos". Over recent years, multiple innovative Contextualized Video interfaces have been proposed to taking advantage of the latest computer graphics and video processing technologies. These interfaces opened a huge design space with numerous design possibilities, each with its own benefits and limitations. To avoid piecemeal understanding of the design space, this dissertation systematically designs and evaluates Contextualized Video interfaces based on a taxonomy of tasks that can potentially benefit from Contextualized Videos. This dissertation first formalizes a design space. New designs are created incrementally along the four major dimensions of the design space. These designs are then empirically compared through a series of controlled experiments using multiple tasks. The tasks are carefully selected from a task taxonomy, which helps to avoid piecemeal understanding of the effect of the designs. Our design practices and empirical evaluations result in a set of design guidelines on how to choose proper designs according to the characteristics of the tasks and the users. Finally, we demonstrate how to apply the design guidelines to prototype a complex interface for a specific video surveillance application. / Ph. D.

The overall goal of this work is to expand the materials library for the fused filament fabrication (FFF) material extrusion additive manufacturing (AM) process through innovations in the FFF process, post-process, and polymer composition. This research was conducted at two opposing ends of the FFF-processing temperature: low processing temperature (<100 °C) for pharmaceutical applications and high processing temperatures (>300 °C) for high-performance structural polymer applications. Both applications lie outside the typical range for FFF (190-260 °C). To achieve these goals, both the material and process were modified. Due to the low processing temperature requirements for pharmaceutical active ingredients, a water-soluble, low melting temperature material (sulfonated poly(ethylene glycol)) series was used to explore how different counterions affect FFF processing. The strong ionic interaction within poly(PEG8k-co-CaSIP) resulted in the best print quality due to the higher viscosity (105 Pa∙s) allowing the material to hold shape in the melt and the high-nucleation producing small spherulites mitigating the layer warping. Fillers were then explored to observe if an ionic filler would produce a similar effect. The ionic filler (calcium chloride) in poly(PEG8k-co-NaSIP) altered the crystallization kinetics, by increasing the nucleation density and viscosity, resulting in improved printability of the semi-crystalline polymer. A methodology for embedding liquids and powders into thin-walled capsules was developed for the incorporation of low-temperature active ingredients into water-soluble materials that uses a higher processing temperature than the actives are compatible with. By tuning the thickness of the printed walls, the time of internal liquid release was controlled during dissolution. This technique was used to enable the release of multiple liquids and powders at different times during dissolution. To enable the printing of high-temperature, high-performance polymers, an inverted desktop-scale heated chamber with the capability of reaching over 300 °C was developed for FFF. The design was integrated onto a FFF machine and was used to successfully print polyphenylsulfone which resulted in a 48% increase in tensile strength (at 200 °C) when compared to printing at room temperature. Finally, the effects of thermal processing conditions for printing ULTEM® 1010 were studied by independently varying the i) nozzle temperature, ii) environment temperature, and iii) post-processing conditions. The nozzle temperature primarily enables flow through the nozzle and needs to be set to at least 360 °C to prevent under extrusion. The environment temperature limits the part warping, as it approaches Tg (217 °C), and improves the layer bonding by decreasing the rate of cooling that allows more time for polymer chain entanglement. Post-processing for a longer time above Tg (18 hrs at 260 °C) promotes further entanglement, which increases the part strength (50% increase in yield strength); however, the part is susceptible to deformation. A post-processing technique was developed to preserve the parts' shape by packing solid parts into powdered salt. / Doctor of Philosophy / Fused filament fabrication (FFF) is the most widely used additive manufacturing (also referred to as 3D printing) process in industry, education, and for hobbyists. However, there is a limited number of materials available for FFF, which limits the potential of using FFF to solve engineering problems. This work focuses on material and machine modifications to enable FFF for use in both pharmaceutical and structural applications. Specifically, many pharmaceutical active ingredients require processing temperatures lower than what FFF typically uses. A low-temperature water-soluble material was altered by incorporating salt ions and ionic fillers separately. The differences in the printability were directly correlated to the measured variations in the viscosity and crystallization material properties. Alternatively, a technique is presented to embed liquids and powders into thin-walled, water-soluble printed parts that are processed using typical FFF temperatures, where the embedded material remains cool. The release time of the embedded material during dissolution is controlled by the thickness of the capsule structure. For structural applications, a machine was developed to allow for the processing of high-performance, high-temperature polymers on a desktop-scale system. This system uses an inverted heated chamber that uses natural convection to be able to heat the air around the part and not the electric components of the machine. The heated environment allows the part to remain at a higher temperature for a longer time, which enables a better bond between printed layers to achieve high-strength printed parts using high-performance materials. This machine was used to characterize the thermal processing effect for printing the high-performance polymer ULTEM® 1010. The nozzle temperature, environment temperature, and post-processing were tested where i) a higher nozzle temperature (360 °C) increases strength and prevents under extrusion, ii) a higher environment temperature (≥200 °C) increases the strength by slowing cooling and decreases warping by limiting the amount of shrinkage the occurs during printing, and iii) post-processing in powdered salt (18 hrs at 260 °C) increases part strength (50%) by allowing the printed roads to fuse.