This paper tackles the challenge of how to compare floor plans with each other. A lot of different methods were used to analyze floor plan images, such as different kinds of pixel-based breadth-first search algorithms for finding walls, doors, and windows. Python-tesseract was used to read text labels in the floor plan, which was of great use when rooms were to be identified. The extracted information from over 1000 floor plans was then used to create a comparison program, which spits out the most similar floor plans to any given floor plan. The results of the extraction part were pretty good for most of the floor plans. Walls, doors, and windows were often accurately found, and the room identification worked very well compared to other known methods. Using the extracted data to find similar floor plans worked splendidly. The extraction part of the project had its flaws and can be improved, but even so, this method of assessing similarity between floor plans works very well. / Den här rapporten tacklar problemet och försöker ge ett svar på hur man kan jämföra planritningar med varandra. Flera olika metoder användes för att analysera planritningar, exempelvis olika typer av pixelbaserade sökalgoritmer för att hitta planritningens väggar, dörrar och fönster. Python-tesseract användes också för att läsa textetiketter i planritningen, vilket var till stor nytta när rummen skulle identifieras. Den extraherade informationen från över 1 000 planritningar användes sedan för att skapa ett jämförelseprogram, som spottar ut de 10 mest liknande planritningarna till en given planritning. Resultatet av extraheringsdelen var väldigt bra för de flesta planritningarna. Se exempelbilden nedan. Väggar, dörrar och fönster hittades ofta korrekt och rumidentifieringen fungerade mycket bra jämfört med andra kända metoder. Att använda den extraherade information för att sedan hitta liknande planlösningar fungerade utmärkt. Extraheringsdelen av projektet hade sina brister och kan förbättras, men trots det fungerar denna metod för att jämföra planlösningar väldigt bra.

The Micro Aerial Vehicless (MAVs) are gaining attention in numerous applications asthese platforms are cheap and can do complex maneuvers. Moreover, most of the commer-cially available MAVs are equipped with a mono-camera. Currently, there is an increasinginterest to deploy autonomous mono-camera MAVs with obstacle avoidance capabilitiesin various complex application areas. Some of the application areas have moving obstaclesas well as stationary, which makes it more challenging for collision avoidance schemes.This master thesis set out to investigate the possibility to avoid moving and station-ary obstacles with a single camera as the only sensor gathering information from thesurrounding environment.One concept to perform autonomous obstacle avoidance is to predict the time near-collision based on a Convolution Neural Network (CNN) architecture that uses the videofeed from a mono-camera. In this way, the heading of the MAV is regulated to maximizethe time to a collision, resulting in the avoidance maneuver. Moreover, another interestingperspective is when due to multiple dynamic obstacles in the environment there aremultiple time predictions for different parts of the Field of View (FoV). The method ismaximizing time to a collision by choosing the part with the largest time to collision.However, this is a complicated task and this thesis provides an overview of it whilediscussing the challenges and possible future directions. One of the main reason was thatthe available data set was not reliable and was not provide enough information for theCNN to produce any acceptable predictions.Moreover, this thesis looks into another approach for avoiding collisions, using objectdetection method You Only Lock Once (YOLO) with the mono-camera video feed. YOLOis a state-of-the-art network that can detect objects and produce bounding boxes in real-time. Because of YOLOs high success rate and speed were it chosen to be used in thisthesis. When YOLO detects an obstacle it is telling where in the image the object is,the obstacle pixel coordinates. By utilizing the images FoV and trigonometry can pixelcoordinates be transformed to an angle, assuming the lens does not distort the image.This position information can then be used to avoid obstacles. The method is evaluated insimulation environment Gazebo and experimental verification with commercial availableMAV Parrot Bebop 2. While the obtained results show the efficiency of the method. To bemore specific, the proposed method is capable to avoid dynamic and stationary obstacles.Future works will be the evaluation of this method in more complex environments with multiple dynamic obstacles for autonomous navigation of a team of MAVs. A video ofthe experiments can be viewed at:https://youtu.be/g_zL6eVqgVM.

Le complexe d'accélérateurs du grand collisionneur de hadrons (le Large Hadron Collider - LHC) sera mis à jour lors du long arrêt de la période LS3 en 2023-2025 pour passer à la phase de haute luminosité (HL-LHC). La luminosité instantanée sera multipliée par 5 pour atteindre 7.5×10³⁴ cm⁻²s⁻¹, ce qui correspond à environ 200 collisions inélastiques par croisement de paquets comparé aux 50 par croisement au LHC. Pendant le fonctionnement du HL-LHC, afin d'atteindre une haute précision dans les études des processus physiques du modèle standard et les recherches de nouvelle physique, le collisionneur à protons devra fournir une luminosité intégrée de l’ordre 400 fb⁻¹ par an pendant une dizaine d’années soit 4000 fb⁻¹ escomptées. Ceci représente un ordre de grandeur supérieur à l'ensemble de la période du LHC. Le détecteur interne (Inner Detector) ATLAS actuel ne sera pas en mesure de faire face efficacement à l'augmentation du taux d'événements et de la dose de rayonnement. Afin d’obtenir des performances au minimum égales ou supérieures à celles de la phase LHC, et tenant compte d’un environnement plus hostile en termes de radiations et d’empilements d’évènements, il a été décidé d’opérer le remplacement complet du trajectomètre interne ou Inner Tracker (ITk); à cette fin, une technologie tout silicium a été choisie. Cette thèse est axée sur l'étude de nouveaux capteurs pixels planaires fins à bords très minces, basés sur le concept d’utilisation de matrices de diodes à implants dopés n sur un substrat dopé p. Ce choix est motivé par les critères de meilleure performance intrinsèque, de radio-tolérance élevée ainsi qu’un coût de production optimisé pour de grandes surfaces. Dans ce travail, des capteurs de différentes épaisseurs allant de 50 μm à 150 μm dotés de bords actifs et minces ont fait l’objet d’études approfondies notamment lors de leur fonctionnement à haut flux de particules chargées. De nombreuses analyses minutieuses ont été menées pour déterminer leur résolution en position à l’aide d’un télescope de faisceau de haute énergie. Les régions d’impacts sur la zone active des pixels ainsi que sur la région des contours ont été scrutées avant et après leur irradiation. En effet, de nombreux résultats obtenus en faisceau de protons et électrons seront montrés, notamment une étude comparative des différents concepts de matrices de capteurs de pixels planaires lus avec la puce de lecture « FE-I4 » en technologie CMOS 130 nm. Préparant la phase future du LHC, nous montrerons les premiers résultats obtenus avec la nouvelle génération de pixels granulaires. Ces matrices ont été couplées à la nouvelle puce de lecture frontale récemment développée au CERN, utilisant la technologie CMOS 65 nm. Ces capteurs dotés d’ une plus fine granularité de 50×50 μm² , ont un pas optimisé lequel est nécessaire pour maintenir un taux d'occupation aussi bas que possible dans un contexte de multiplicités de particules chargées très élevées. Dans ce travail, une contribution personnelle à l’électronique de lecture sera détaillée, en particulier les études ont été menées sur la puce nommée «Ring-Oscillator» ou moniteur de radiations, développée au laboratoire. On décrira son comportement dynamique en fonction de la température, de la tension nominale et en conditions hautement radiatives (500 MRad). La mise au point d’un nouvel outil de caractérisation de détecteurs pixels a fait l’objet d’un développement important. Grâce à un ensemble basé sur une excitation laser de 1060 nm, il sera possible de disposer d’un système précis et autonome capable de mesurer rapidement les caractéristiques fonctionnelles des matrices de pixels avec une excellente résolution spatiale. Les caractéristiques de cet outil feront l’objet d’une présentation exhaustive. / The Large Hadron Collider (LHC) will go through the accelerator complex upgrade during the LS3 long shutdown in 2023-2025 to move to the High Luminosity phase (HL-LHC). As a result, an instantaneous luminosity will increase sevenfold to 7.5×10³⁴ cm⁻²s⁻¹, corresponding to approximately 200 inelastic collisions per bunch-crossing, whereas the LHC runs resulted in up to 50 collisions per bunch-crossing. During the operation of the HL-LHC, in order to achieve high-precision in studies of Standard Model processes and searches for new physics, about 4000 fb⁻¹ of integrated luminosity be collected, which is of an order of magnitude larger than over the entire LHC period. The present ATLAS Inner Detector (ID) will not be able to efficiently cope with the increased event rate and radiation dose. Due to this fact the complete replacement of the ID is foreseen with fully silicon Inner Tracker (ITk) to provide high tracking performance in harsher environment delivered by the HL-LHC. This thesis is focused on the study of new n+-in-p planar silicon sensors, as a promising option to instrument the ITk pixel layers, considering their radiation hardness and cost-effectiveness. Sensors of different thicknesses ranging from 50 µm to 150 µm of active and slim edge designs have been tested at a high energy particle beam to investigate hit efficiency performance analyzed on the pixel active area and on the edge area before and after irradiation. The test beam results and their comparison for the different designs of the pixel sensors compatible with FE-I4 readout chip are discussed. Also, the first results on test beam characterization of the pixel modules employing a newly developed prototype of readout chip for the ITk, RD53A chip, implemented in 65 nm CMOS technology, were obtained. The sensors with the decreased to 25×100 µm² and 50×50 µm² pixel pitch to maintain the lower level of occupancy at high particle multiplicity were measured. Additionally, the tests of ring oscillators, contained in RD53A chip, which may be used as a monitor of the received radiation dose, were carried out depending on temperature, supplied voltage and irradiation level up to 500 MRad. Finally, the test bench setup for silicon pixel detectors characterization using an infra-red laser has been the subject of an original development in this thesis. The setup was developed in the clean room at Laboratoire de l'accélérateur linéaire (LAL) and it is capable of rapidly measuring the functional characteristics, providing a flexible charge injection with well-defined hit position to characterize the silicon pixel matrixes. The software to control the setup was created using LabVIEW programming environment. The results of the measurements with the FE-I4 module implemented with openings allowing the laser beam passage on a sensor backside are presented in this thesis.

Aim of this project is to evaluate effectivity of various texture features within the context of image processing, particulary the task of texture recognition and classification. My work focuses on comparing and discussion of usage and efficiency of texture features based on local binary patterns and co- ccurence matrices. As classification algorithm is concerned, cluster analysis was choosen.

[pt] Redes neurais profundas são responsáveis pelo grande progresso em diversas tarefas perceptuais, especialmente nos campos da visão computacional,reconhecimento de fala e processamento de linguagem natural. Estes resultados produziram uma mudança de paradigma nas técnicas de reconhecimentode padrões, deslocando a demanda do design de extratores de característicaspara o design de arquiteturas de redes neurais. No entanto, o design de novas arquiteturas de redes neurais profundas é bastante demandanteem termos de tempo e depende fortemente da intuição e conhecimento de especialistas,além de se basear em um processo de tentativa e erro. Neste contexto, a idea de automatizar o design de arquiteturas de redes neurais profundas tem ganhado popularidade, estabelecendo o campo da busca por arquiteturas neurais(NAS - Neural Architecture Search). Para resolver o problema de NAS, autores propuseram diversas abordagens envolvendo o espaço de buscas, a estratégia de buscas e técnicas para mitigar o consumo de recursos destes algoritmos. O Q-NAS (Quantum-inspired Neural Architecture Search) é uma abordagem proposta para endereçar o problema de NAS utilizando um algoritmo evolucionário com inspiração quântica como estratégia de buscas. Este método foi aplicado de forma bem sucedida em classificação de imagens, superando resultados de arquiteturas de design manual nos conjuntos de dados CIFAR-10 e CIFAR-100 além de uma aplicação de mundo real na área da sísmica. Motivados por este sucesso, propõe-se nesta Dissertação o SegQNAS (Quantum-inspired Neural Architecture Search applied to Semantic Segmentation), uma adaptação do Q-NAS para a tarefa de segmentação semântica. Diversos experimentos foram realizados com objetivo de verificar a aplicabilidade do SegQNAS em dois conjuntos de dados do desafio Medical Segmentation Decathlon. O SegQNAS foi capaz de alcançar um coeficiente de similaridade dice de 0.9583 no conjunto de dados de baço, superando os resultados de arquiteturas tradicionais como U-Net e ResU-Net e atingindo resultados comparáveis a outros trabalhos que aplicaram NAS a este conjunto de dados, mas encontrando arquiteturas com muito menos parãmetros. No conjunto de dados de próstata, o SegQNAS alcançou um coeficiente de similaridade dice de 0.6887 superando a U-Net, ResU-Net e o trabalho na área de NAS que utilizamos como comparação. / [en] Deep neural networks are responsible for great progress in performance for several perceptual tasks, especially in the fields of computer vision, speech recognition, and natural language processing. These results produced a paradigm shift in pattern recognition techniques, shifting the demand from feature extractor design to neural architecture design. However, designing novel deep neural network architectures is very time-consuming and heavily relies on experts intuition, knowledge, and a trial and error process. In that context, the idea of automating the architecture design of deep neural networks has gained popularity, establishing the field of neural architecture search (NAS). To tackle the problem of NAS, authors have proposed several approaches regarding the search space definition, algorithms for the search strategy, and techniques to mitigate the resource consumption of those algorithms. Q-NAS (Quantum-inspired Neural Architecture Search) is one proposed approach to address the NAS problem using a quantum-inspired evolutionary algorithm as the search strategy. That method has been successfully applied to image classification, outperforming handcrafted models on the CIFAR-10 and CIFAR-100 datasets and also on a real-world seismic application. Motivated by this success, we propose SegQNAS (Quantum-inspired Neural Architecture Search applied to Semantic Segmentation), which is an adaptation of Q-NAS applied to semantic segmentation. We carried out several experiments to verify the applicability of SegQNAS on two datasets from the Medical Segmentation Decathlon challenge. SegQNAS was able to achieve a 0.9583 dice similarity coefficient on the spleen dataset, outperforming traditional architectures like U-Net and ResU-Net and comparable results with a similar NAS work from the literature but with fewer parameters network. On the prostate dataset, SegQNAS achieved a 0.6887 dice similarity coefficient, also outperforming U-Net, ResU-Net, and outperforming a similar NAS work from the literature.

Tesis por compendio / [ES] Las playas son ambientes ecológicos sumamente valiosos donde a lo largo de una frágil franja de transición converge el entorno terrestre y el medio marino. Durante el último siglo, la mejora en la comprensión de los procesos físicos que ocurren en la zona costera se ha convertido en un asunto de máxima importancia. Para abordar una planificación coherente de la gestión costera se requiere tomar en consideración el dinamismo de los diferentes cambios morfológicos que caracterizan estos ambientes a distintas escalas espaciales y temporales. El límite tierra-agua varía en función de la posición del nivel del mar y de la forma del perfil de playa que continuamente queda modelado por las olas incidentes. Intentar modelizar la respuesta de un paisaje tan voluble geomorfológicamente como las playas requiere disponer de múltiples medidas registradas con suficiente precisión para poder reconocer su respuesta frente a la acción de los distintos agentes geomórficos. Para ello resulta esencial disponer de diferentes sistemas de monitorización capaces de registrar de forma sistemática la línea de costa con exactitud y efectividad. Se requieren nuevos métodos y herramientas informáticas que permitan capturar, caracterizar y analizar eficientemente la información con el objeto de obtener indicadores con significación geomorfológica de calidad. En esto radica el objetivo de la presente tesis doctoral, centrándose en el desarrollo de herramientas y procedimientos eficientes para la monitorización costera mediante el uso de imágenes satelitales y fotografías terrestres. El trabajo aporta soluciones de procesamiento de imágenes de satélite y fotogramétricas a científicos, ingenieros y gestores costeros, proporcionando resultados que evidencian la gran utilidad de estas técnicas viables y de bajo coste para la monitorización costera. Mediante ellas se puede convertir información pública existente y de libre acceso (imágenes satelitales, datos de video cámaras o fotografías de la ciudadanía) en datos de alta calidad para el monitoreo de los cambios morfológicos de las playas, y lograr así una consiguiente gestión sostenible de los recursos costeros. / [CA] Les platges són ambients ecològics summament valuosos on al llarg d'una feble franja de transició convergeix l'entorn terrestre i el medi marí. En l'últim segle, la millora en la comprensió dels processos físics que ocorren en la zona costanera s'ha convertit en un assumpte de màxima importància. Per a abordar una planificació coherent de la gestió costanera es requereix prendre en consideració el dinamisme dels diferents canvis morfològics que caracteritzen aquests ambients a diferents escales espacials i temporals. El límit terra-aigua varia en funció de la posició del nivell del mar i de la forma del perfil de platja que contínuament queda modelat per les ones incidents. Intentar modelitzar la resposta d'un paisatge tan voluble geomorfològicament com les platges requereix disposar de múltiples mesures registrades amb suficient precisió per poder reconèixer la seua resposta enfront de l'acció dels diferents agents geomòrfics. Per tant, resulta essencial disposar de diferents sistemes de monitoratge capaços de registrar de forma sistemàtica la línia de costa amb exactitud i efectivitat. Es requereixen nous mètodes i eines informàtiques que permeten capturar, caracteritzar i analitzar eficientment la informació a fi d'obtindre indicadors amb significació geomorfològica de qualitat. En això radica l'objectiu de la present tesi doctoral, que es centra en el desenvolupament d'eines i procediments eficients per al monitoratge costaner mitjançant l'ús d'imatges de satèl·lit i fotografies terrestres. El treball aporta solucions de processament d'imatges de satèl·lit i fotogramètriques a científics, enginyers, polítics i gestors costaners, proporcionant resultats que evidencien la gran utilitat d'aquestes tècniques factibles i de baix cost per a la monitorització costanera. Mitjançant aquestes es pot convertir informació pública existent i de lliure accés (imatges de satèl·lit, dades de videocàmeres o fotografies de la ciutadania) en dades d'alta qualitat per al monitoratge dels canvis morfològics de les platges, i aconseguir així una consegüent gestió sostenible dels recursos costaners. / [EN] Beaches are extremely valuable ecological spaces where terrestrial and marine environments converge along a fragile transition strip. An improvement in our understanding of the physical processes that occur in the coastal zone has become increasingly important during the last century. To approach a coherent planning of coastal management it is necessary to consider the dynamism of the various morphological changes that characterize these environments at different spatial and temporal scales. The land-water boundary varies according to the sea level and the shape of a beach profile that is continuously modelled by incident waves. Attempting to model the response of a landscape as geomorphologically volatile as beaches requires multiple precise measurements to recognize responses to the actions of various geomorphic agents. It is therefore essential to have monitoring systems capable of systematically recording the shoreline accurately and effectively. New methods and tools are required to efficiently capture, characterize, and analyze information - and so obtain geomorphologically significant indicators. This is the aim of the doctoral thesis, focusing on the development of tools and procedures for coastal monitoring using satellite images and terrestrial photographs. The work brings satellite image processing and photogrammetric solutions to scientists, engineers, and coastal managers by providing results that demonstrate the usefulness of these viable and low-cost techniques. Existing and freely accessible public information (satellite images, video-derived data, or crowd-sourced photographs) can be converted into high quality data for monitoring morphological changes on beaches and thus help achieve a sustainable management of coastal resources. / Agradecer al Ministerio de Educación, Cultura y Deporte del Gobierno de España por la beca predoctoral FPU, y por las ayudas de movilidad concedidas, que han permitido que esta Tesis Doctoral fuera una realidad. También a los proyectos AICO/2015/098 y CGL2015-69906-R financiados respectivamente por la Generalitat Valenciana y por el Ministerio de Economía y Competitividad. / Sánchez García, E. (2019). Photogrammetry and image processing techniques for beach monitoring [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/123956 / TESIS / Compendio

Technology has revamped every aspect of our life, one of those various facets is fashion industry. Plenty of deep learning architectures are taking shape to augment fashion experiences for everyone. There are numerous possibilities of enhancing the fashion technology with deep learning. One of the key ideas is to generate fashion style and recommendation using artificial intelligence. Likewise, another significant feature is to gather reliable information of fashion trends, which includes analysis of existing fashion related images and data. When specifically dealing with images, localisation and segmentation are well known to address in-depth study relating to pixels, objects and labels present in the image. In this master thesis a complete framework is presented to perform localisation and segmentation on fashionista images. This work is a part of an interesting research work related to Fashion Style detection and Recommendation. Developed solution aims to leverage the possibility of localising fashion items in an image by drawing bounding boxes and labelling them. Along with that, it also provides pixel-wise semantic segmentation functionality which extracts fashion item label-pixel data. Collected data can serve as ground truth as well as training data for the aimed deep learning architecture. A study related to localisation and segmentation of videos has also been presented in this work. The developed system has been evaluated in terms of flexibility, output quality and reliability as compared to similar platforms. It has proven to be fully functional solution capable of providing essential localisation and segmentation services while keeping the core architecture simple and extensible. / Tekniken har förnyat alla aspekter av vårt liv, en av de olika fasetterna är modeindustrin. Massor av djupa inlärningsarkitekturer tar form för att öka modeupplevelser för alla. Det finns många möjligheter att förbättra modetekniken med djup inlärning. En av de viktigaste idéerna är att skapa modestil och rekommendation med hjälp av artificiell intelligens. På samma sätt är en annan viktig egenskap att samla pålitlig information om modetrender, vilket inkluderar analys av befintliga moderelaterade bilder och data. När det specifikt handlar om bilder är lokalisering och segmentering väl kända för att ta itu med en djupgående studie om pixlar, objekt och etiketter som finns i bilden. I denna masterprojekt presenteras en komplett ram för att utföra lokalisering och segmentering på fashionista bilder. Detta arbete är en del av ett intressant forskningsarbete relaterat till Fashion Style detektering och rekommendation. Utvecklad lösning syftar till att utnyttja möjligheten att lokalisera modeartiklar i en bild genom att rita avgränsande lådor och märka dem. Tillsammans med det tillhandahåller det även pixel-wise semantisk segmenteringsfunktionalitet som extraherar dataelementetikett-pixeldata. Samlad data kan fungera som grundsannelse samt träningsdata för den riktade djuplärarkitekturen. En studie relaterad till lokalisering och segmentering av videor har också presenterats i detta arbete. Det utvecklade systemet har utvärderats med avseende på flexibilitet, utskriftskvalitet och tillförlitlighet jämfört med liknande plattformar. Det har visat sig vara en fullt fungerande lösning som kan tillhandahålla viktiga lokaliseringsoch segmenteringstjänster samtidigt som kärnarkitekturen är enkel och utvidgbar.

This essay examines the ways the decorated farmhouses known as hälsingegårdar of Hälsingland, Sweden, are depicted in art from the late nineteenth century Düsseldorf school of painting to contemporary pixel art and how the ways of depiction have changed over time. Three different works of art are analyzed with Erwin Panofsky’s model for iconographic and iconological analysis. Results show a gradual difference in technique and the artist's personal attitudes towards the hälsingegårdar. Starting with the genre painting’s idyllic interpretation of the peasant life as the height of nationalist values to the more provincialist cherry-picking of the late twentieth century naive painting and the final contemporary pixel art piece of ambivalence and distancing from the hälsingegårdar as a means of interpretation and understanding.

The intensification of aridity due to anthropogenic climate change is likely to have a large impact on the growth and survival of plant species in the southwestern U.S. where species are already vulnerable to high temperatures and limited precipitation. Global climate change impacts plants through a rising temperature effect, CO2 effect, and land management. In order to forecast the impacts of global climate change, it is necessary to know the current conditions and create a baseline for future comparisons and to understand the factors and players that will affect what happens in the future. The objective of Chapter 1 is to create the very first high resolution, accurate, park-wide map that shows the distribution of dominant plants on the Colorado Plateau and serves as a baseline for future comparisons of species distribution. If we are going to forecast what species have already been impacted by global change or will likely be impacted in the future, we need to know their physiology. Chapter 2 surveys the physiology of the twelve most abundant non-tree species on the Colorado Plateau to help us forecast what climate change might do and to understand what has likely already occurred. Chapter 1. Our objective was to create an accurate species-level classification map using a combination of multispectral data from the World View-3 satellite and hyperspectral data from a handheld radiometer to compare pixel-based and object-based classification. We found that overall, both methods were successful in creating an accurate landscape map. Different functional types could be classified with fairly good accuracy in a pixel-based classification but to get more accurate species-level classification, object-based methods were more effective (0.915, kappa coefficient=0.905) than pixel-based classification (0.79, kappa coefficient=0.766). Although spectral reflectance values were important in classification, the addition of other features such as brightness, texture, number of pixels, size, shape, compactness, and asymmetry improved classification accuracy.Chapter 2. We sought to understand if patterns of gas exchange to changes in temperature and CO2 can explain why C3 shrubs are increasing, and C3 and C4 grasses are decreasing in the southwestern U.S. We conducted seasonal, leaf-level gas exchange surveys, and measured temperature response curves and A-Ci response curves of common shrub, forb, and grass species in perennial grassland ecosystems over the year. We found that the functional trait of being evergreen is increasingly more successful in climate changing conditions with warmer winter months. Grass species in our study did not differentiate by photosynthetic pathway; they were physiologically the same in all of our measurements. Increasing shrub species, Ephedra viridis and Coleogyne ramosissima displayed functional similarities in response to increasing temperature and CO2.

The intensification of aridity due to anthropogenic climate change is likely to have a large impact on the growth and survival of plant species in the southwestern U.S. where species are already vulnerable to high temperatures and limited precipitation. Global climate change impacts plants through a rising temperature effect, CO2 effect, and land management. In order to forecast the impacts of global climate change, it is necessary to know the current conditions and create a baseline for future comparisons and to understand the factors and players that will affect what happens in the future. The objective of Chapter 1 is to create the very first high resolution, accurate, park-wide map that shows the distribution of dominant plants on the Colorado Plateau and serves as a baseline for future comparisons of species distribution. If we are going to forecast what species have already been impacted by global change or will likely be impacted in the future, we need to know their physiology. Chapter 2 surveys the physiology of the twelve most abundant non-tree species on the Colorado Plateau to help us forecast what climate change might do and to understand what has likely already occurred. Chapter 1. Our objective was to create an accurate species-level classification map using a combination of multispectral data from the World View-3 satellite and hyperspectral data from a handheld radiometer to compare pixel-based and object-based classification. We found that overall, both methods were successful in creating an accurate landscape map. Different functional types could be classified with fairly good accuracy in a pixel-based classification but to get more accurate species-level classification, object-based methods were more effective (0.915, kappa coefficient=0.905) than pixel-based classification (0.79, kappa coefficient=0.766). Although spectral reflectance values were important in classification, the addition of other features such as brightness, texture, number of pixels, size, shape, compactness, and asymmetry improved classification accuracy.Chapter 2. We sought to understand if patterns of gas exchange to changes in temperature and CO2 can explain why C3 shrubs are increasing, and C3 and C4 grasses are decreasing in the southwestern U.S. We conducted seasonal, leaf-level gas exchange surveys, and measured temperature response curves and A-Ci response curves of common shrub, forb, and grass species in perennial grassland ecosystems over the year. We found that the functional trait of being evergreen is increasingly more successful in climate changing conditions with warmer winter months. Grass species in our study did not differentiate by photosynthetic pathway; they were physiologically the same in all of our measurements. Increasing shrub species, Ephedra viridis and Coleogyne ramosissima displayed functional similarities in response to increasing temperature and CO2.