Um dos desafios da Inteligência Artificial aplicada em jogos é o aprendizado de comportamento, em que o objetivo é utilizar estatísticas obtidas da interação entre jogador e jogo de modo a reconhecer características particulares de um jogador ou monitorar a evolução de seu comportamento no decorrer do tempo. A maior parte dos trabalhos na área emprega modelos previamente aprendidos, por meio da utilização de algoritmos de Aprendizado de Máquina. Entretanto, são poucos os trabalhos que consideram que o comportamento de um jogador pode evoluir no tempo e que, portanto, reconhecer quando essas mudanças ocorrem é o primeiro passo para produzir jogos que se adaptam automaticamente às capacidades do jogador. Para detectar variações comportamentais em um jogador, são necessários algoritmos que processem dados de modo incremental. Esse pré-requisito motiva o estudo de algoritmos para detecção de mudanças da área de Mineração em Fluxos Contínuos de Dados. Entretanto, algumas das características dos algoritmos disponíveis na literatura inviabilizam sua aplicação direta ao problema de detecção de mudança em jogos. Visando contornar essas dificuldades, esta tese propõe duas novas abordagens para detecção de mudanças de comportamento. A primeira abordagem é baseada em um algoritmo incremental de agrupamento e detecção de novidades que é independente do número e formato dos grupos presentes nos dados e que utiliza um mecanismo de janela deslizante para detecção de mudanças de comportamento. A segunda abordagem, por outro lado, é baseada na comparação de janelas de tempo consecutivas utilizando espectrogramas gerados a partir dos dados contidos em cada janela. Os resultados experimentais utilizando simulações e dados de jogos comerciais indicam a aplicabilidade dos algoritmos propostos na tarefa de detecção de mudanças de comportamento de um jogador, assim como mostram sua vantagem em relação a outros algoritmos para detecção de mudança disponíveis na literatura / One of the challenges of Artificial Intelligence applied to games is behavior learning, where the objective is to use statistics derived from the interaction between the player and the game environment in order to recognize particular player characteristics or to monitor the evolution of a players behavior along time. The majority of work developed in this area applies models that were previously learned through the use of Machine Learning techniques. However, only a few pieces of work consider that the players behavior can evolve over time and, therefore, recognizing when behavior changes happen is the first step towards the production of games that adapt to the players needs. In order to detect changes in the behavior of a player, incremental algorithms are necessary, what motivates the study of change detection algorithms from the area of Data Stream Mining. However, some of the characteristics of the algorithms available in the literature make their application to the task of change detection in games unfeasible. To overcome these difficulties, this work proposes two new approaches for change detection. The first approach is based on an incremental clustering and novelty detection algorithm which is independent of the number and format of clusters and uses a mechanism for change detection based on sliding windows. The second approach, on the other hand, is based on the comparison of consecutive time windows using spectrograms created from the data inside each window. Experimental results using simulations and data from commercial games indicate the applicability of the proposed algorithms in the task of detecting a players changing behavior, as well as present their advantage when compared to other change detection algorithms available in the literature

Nesta dissertação está proposta uma metodologia para fins de detecção de mudanças do uso do solo em imagens multitemporais de sensoriamento remoto. Em lugar de classificar os pixels de imagens que cobrem uma cena, em duas classes exaustivas e mutuamente excludentes (mudança, não-mudança), propõe-se adotar uma abordagem do tipo fuzzy, na qual são estimados os graus de pertinência às classes mudança e não-mudança. Com este objetivo adota-se aqui uma abordagem em nível de sub-pixel na estimação dos graus de pertinência para cada pixel. Esta abordagem se mostra mais adequada para fins de modelagem do que ocorre em cenas naturais, onde as alterações que acontecem ao longo de um período de tempo tendem a apresentar uma variação contínua em lugar de discreta. Em uma segunda etapa, os graus de pertinência estimados recebem um ajustamento adicional por meio da introdução de informações de contexto espacial. A metodologia proposta foi testada por meio de três experimentos, um empregando uma imagem sintética e dois utilizando imagens reais. A partir da análise quantitativa dos resultados e comparação com estudos semelhantes, comprova-se a adequação da metodologia proposta. / In this dissertation it is proposed a new methodology to land use change detection in remote sensing multitemporal image data. Rather than applying a rigid labeling of the pixels in the image data into two classes (change, no-change), we propose estimating the degrees of membership to classes change and no-change in a fuzzy-like fashion. To this end, a sub-pixel approach is implemented to detect the degree of change in every pixel. This methodology aims at modeling natural scenes in a more realistic way, since changes in natural scenes tend to occur in a continuum rather than in a sharp distinctive way. In a second step, the estimated values for the degrees of membership are further refined by means of spatial context information. Three experiments were performed to test the proposed methodology, one employing synthetic data and two using real image data. From the quantitative analysis of the results and from similar studies we can prove the adequacy of the proposed methodology.

A detecção de mudanças na superfície terrestre é o principal objetivo em aplicações de sensoriamento remoto multitemporal. Sabe-se que imagens adquiridas em datas distintas tendem a ser altamente influenciadas por problemas radiométricos e de registro. Utilizando imagens de fração, obtidas a partir do modelo linear de mistura espectral (MLME), problemas radiométricos podem ser minimizados e a interpretação dos tipos de mudança na superfície terrestre é facilitada, pois as frações têm um significado físico direto. Além disso, interpretações ao nível de subpixel são possíveis. Esta tese propõe três algoritmos – rígido, suave e fuzzy – para a detecção de mudanças entre um par de imagens de fração, gerando mapas de mudança como produtos finais. As propostas requerem a suposição de normalidade multivariada para as diferenças de fração e necessitam de pouca intervenção por parte do analista. A proposta rígida cria mapas de mudança binários seguindo a mesma metodologia de um teste de hipóteses, baseando-se no fato de que os contornos de densidade constante na distribuição normal multivariada são definidos por valores da distribuição qui-quadrado, de acordo com a escolha do nível de confiança. O classificador suave permite gerar estimativas da probabilidade do pixel pertencer à classe de mudança, a partir de um modelo de regressão logística. Essas probabilidades são usadas para criar um mapa de probabilidades de mudança. A abordagem fuzzy é aquela que melhor se adapta ao conceito de pixel mistura, visto que as mudanças no uso e cobertura do solo podem ocorrer em nível de subpixel. Com base nisso, mapas dos graus de pertinência à classe de mudança foram criados. Outras ferramentas matemáticas e estatísticas foram utilizadas, tais como operações morfológicas, curvas ROC e algoritmos de clustering. As três propostas foram testadas utilizando-se imagens sintéticas e reais (Landsat-TM) e avaliadas qualitativa e quantitativamente. Os resultados indicam a viabilidade da utilização de imagens de fração em estudos de detecção de mudanças por meio dos algoritmos propostos. / Land cover change detection is a major goal in multitemporal remote sensing applications. It is well known that images acquired on different dates tend to be highly influenced by radiometric differences and registration problems. Using fraction images, obtained from the linear model of spectral mixing (LMSM), radiometric problems can be minimized and the interpretation of changes in land cover is facilitated because the fractions have a physical meaning. Furthermore, interpretations at the subpixel level are possible. This thesis presents three algorithms – hard, soft and fuzzy – for detecting changes between a pair of fraction images. The algorithms require multivariate normality for the differences among fractions and very little intervention by the analyst. The hard algorithm creates binary change maps following the same methodology of hypothesis testing, based on the fact that the contours of constant density are defined by chi-square values, according to the choice of the probability level. The soft one allows for the generation of estimates of the probability of each pixel belonging to the change class by using a logistic regression model. These probabilities are used to create a map of change probabilities. The fuzzy approach is the one that best fits the concept behind the fraction images because the changes in land cover can occurr at a subpixel level. Based on these algorithms, maps of membership degrees were created. Other mathematical and statistical techniques were also used, such as morphological operations, ROC curves and a clustering algorithm. The algorithms were tested using synthetic and real images (Landsat-TM) and the results were analyzed qualitatively and quantitatively. The results indicate that fraction images can be used in change detection studies by using the proposed algorithms.

Karlstads vision är att bli en kommun med 100 000 invånare till år 2031. För att kommunen ska nå målet innebär det att fler bostäder behöver byggas. Det innebär att det kommer bli markanvändningsförändringar inom Karlstadskommun. Den urbana miljön bör påverkats när antalet invånare i Karlstad ökar. Fjärranalys är ett sätt att kunna inventera jordytans biofysiska egenskaper och antropogena förändringar. Fjärranalysen har även använts för att kvantifiera och kartlägga ekosystem-egenskaper. Ekosystemtjänster har positiv påverkan på människor, eftersom de hjälper till med att minska stress, ångest, och har positiv inverkan på återhämtning. Medvetenheten av ekosystemtjänster har ökat med åren och det medför att besluttagares och allmänheten är mer medvetna på värdet av de varor och tjänster som ekosystemtjänsterna bidrar till. Att uttrycka värdena av ekosystemtjänster i pengarvärde är ett viktigt verktyg för att ytterligare öka medvetenheten och betydelsen av ekosystem och mångfald till beslutsfattare. Med hjälp av fjärranalys går det att upptäcka att det har skett en del marktäckes- och markanvändningsförändringar runt om i Karlstads tätort. Tydliga förändringar upptäcktes runt Välsviken och vid Bergviks köpcenter, har det uppkommit nya byggnader. Vid den gamla flygplatsen har ett nytt bostadskvarter byggts samt vid områdena Stockfallet och Campus har expanderats med fler bostäder och andra byggnader.Studien har visat att nybyggnationer sker på bekostnad av främst skog. Detta medför att en del av de naturliga ekosystemtjänsterna försvinner. Totalt har 1 016 ha skog och 154 ha vatten försvunnit, det har tillkommit 196 ha öppen mark och bebyggelse har ökat med 975 ha i Karlstads tätort.Ett ekosystemvärde på 58 986 Int$/ha/år för skog har försvunnit från Karlstads tätortsområde, det motsvarar en förminskning på 30 % på 13 år. Medans öppen mark har ett ekosystemvärde som motsvarar 6 272 Int$/ha/år som har tillkommit i området. Det är en ökning på 0,7 %. Vattnet har ett värde på 2 310 Int$/ha/år, som motsvarar en förminskning på 0,6 %. Ett ekosystemtjänstvärde av totalt 61 238 Int$/ha/år har försvunnit från Karlstad. Invånare som bor i centrum kommer få längre till ekosystemtjänsterna som finns i skogen, men även mycket av de natursköna vyerna kommer försvinna från staden. Eftersom växlighet ingår i infrastrukturen kommer invånarna fortfarande ha tillgång till en del ekosystemtjänster i staden. / The municipality of Karlstad has a vision to reach 100,000 inhabitants by year 2031, which leads to that more housing is needed. In conclusion, more housing leads to land cover changes in the municipality. Remote sensing is one way to discoverer soil biophysical properties and anthropogenic changes. It has even been used to quantify and map ecosystem properties. Ecosystem services have a positive effect on people, because they help to reduce stress, depression and have a good impact on recovery. The awareness of ecosystem services has increased, it means that decision makers and the public are now more aware of the significant value of ecosystem services. Through remote sensing, land use and land cover changes can be observed in Karlstad’s urban area. As prominent changes, in the areas around Välsviken and Bergviks shopping mall, new buildings were created. At the old airport, a new residential area was constructed. Further changes could be observed in areas around Stockfallet and Campus where more residential buildings were constructed. When forests are converted to new residential areas, some natural ecosystem services disappear. The municipality of Karlstad has experienced losses that amount to a total of 1017 ha forest and 154 ha water in 13 years. It has an ecosystem service value of 61 238 Int$/ha/year and that has disappeared from Karlstad. It corresponds a loss of 30 percent of the ecosystem services. While open fields have increased with 6272 ha and has a value of 6272 Int$/ha/year. Citizens that are living in the centre is going to have a longer distance to ecosystem services in the forest. Much of the scenic views will disappeared from Karlstad. Because vegetation is included in infrastructures will residents still have access to ecosystem services in the city.

Le changement d’usage des terres, liées à l’agriculture et à la foresterie, engendre une perte importante de biodiversité et représente une part importante de nos émissions de GES à l’origine du changement climatique. Le mécanisme de Réduction des Émissions de la Déforestation et Dégradation des forêts, conservation, gestion durable et restauration des stocks de carbone (REDD+) initié il y a dix ans. peine à se mettre en place du fait de nombreuses contraintes politiques et scientifiques. Malgré l’existence de lignes directrices élaborées par la communauté scientifique internationale, des outils et données sont nécessaires afin de fournir des informations précises, à moindre coût et utilisables à différentes échelles. L’objectif de cette thèse est de développer des méthodologies innovantes pour réduire les incertitudes sur les estimations des émissions et séquestrations de CO2 associées à la déforestation, dégradation et régénération des terres. Madagascar, pays engagé dans le REDD+ depuis huit ans et soumis à des pertes importantes de biodiversité et de couvert forestier est pris comme exemple. Trois études complémentaires ont été réalisées : i) le suivi de la déforestation en région tropical humide et sec par satellites, ii) l’estimation des stocks de carbone dans les sols et les forêts et iii) la modélisation des changements d’usage de terres. Nous avons développé une nouvelle méthodologie de suivi de la déforestation à Madagascar permettant de tenir compte de la définition des forêts et améliorer la prise en compte des petites parcelles de défriche brulis. Les chiffres de la déforestation, variant d’une région à une autre, ont ainsi été actualisés jusqu’en 2013. Une méthodologie innovante de cartographie des stocks de carbone dans le sol à des résolutions fines et à des échelles régionales a été mise au point en couplant de nombreux facteurs environnementaux et un inventaire de terrain à l’aide d’un modèle d’apprentissage automatique. Ce modèle spatial du carbone a été appliqué sur des images satellites acquises vingt année plus tôt afin d’évaluer la dégradation des stocks de carbone du sol et leur régénération potentielle. Des facteurs de perte et gains de carbone dans le sol ont pu ainsi être estimés. Enfin, une approche de modélisation des changements d’usage des terres a permis de mieux comprendre les facteurs biophysiques et socio-économiques liées à la déforestation, dégradation des terres et régénération, et proposer des scénarios spatialisés pour aider les décideurs. Les résultats obtenus dans cette thèse et les méthodologies développées permettent d’alimenter les discussions et documents concernant la stratégie REDD+ de Madagascar. Elle contribue et vise à une meilleure gestion des agro-écosystèmes par la fourniture d’informations spatiales justes, précises spatialement et pertinentes à grande échelle. / Land use change due to agriculture and forestry, generates a significant loss of biodiversity and is an important part of our greenhouse gas (GHG) emissions causing climate change. The Reduction of Emissions from Deforestation and Forest Degradation, conservation, sustainable management and restoration of carbon stocks (REDD+) mechanism initiated ten years ago is struggling to establish because of many political and scientific constraints. Despite the existence of guidelines developed by the international scientific community, tools and data necessary to provide accurate, cost and usable at different scales. The objective of this thesis is to develop innovative methods to reduce uncertainties in the estimates of CO2 emissions and sequestrations from deforestation, degradation and land regeneration. Madagascar, a country committed in REDD+ for eight years and subjected to significant losses of biodiversity and forest cover, is taken as an example. Three complementary studies were carried out: i) monitoring of deforestation in tropical humid and dry regions, ii) estimates of carbon stocks in soils and forests and iii) land use change model. We have developed a new methodology for monitoring deforestation in Madagascar considering the national definition of forests and accounted for small plots of slash and burn practices. The figures of deforestation vary from one region to another, and have been updated to 2013. An innovative methodology for soil organic carbon stock mapping at fine resolution and regional scale has been developed by coupling many environmental factors and a field inventory using a machine learning model. This spatial carbon model was applied on satellite images acquired twenty year ago to assess the degradation of soil carbon stocks and potential regeneration. Loss and gain factors due to various land use change were estimated. Finally, the land use change framework developed allowed us to understand the biophysical and socio-economic factors related to deforestation, land degradation and regeneration, and provide spatially scenarios to assist policy makers. The results obtained in this thesis and the methodologies developed allow to feed the discussions and documents relating to the REDD + strategy in Madagascar. It contributes and is aimed at a better management of agro-ecosystems by providing accurate spatial information, locally relevant and globally consistent.

Den 8:e januari 2005 inträffade en av de mest förödande stormarna i Sveriges historia då hundratusentals blev strömlösa och sju personer miste livet. Stormen Gudrun drabbade centrala Götaland värst och uppemot nio årsavverkningar skog beräknas ha fällts i vissa områden. Tidigare studier av stormen har genomförts på uppdrag av Skogsstyrelsen där resultaten visar att andel stormfälld skogsmarksareal var 11 % i värst drabbade Ljungby kommun, och ca 80 % av all den stormfällda skogen var gran, 18 % var tall och 2 % var löv. Syftet med arbetet är att undersöka mängden stormfälld skog efter stormen Gudrun genom analys av satellitburen fjärranalysdata. Även andelen stormfälld barr- och lövskog beräknades och resultaten jämfördes med de rapporter skrivna för Skogsstyrelsen. Även andelen stormfälld skog som är återbeskogad år 2016 beräknades. En förändringsanalys med satellitbilder från Landsat 5, tagna åren 2004 och 2005, genomfördes vilken inkluderade en skogsmask som skapades genom övervakad MLC-klassificering. Skogsmasken användes för att utesluta ointressanta områden i analyserna. Resultatet användes sedan för analys av andelen stormfälld barr- och lövskog samt för analys av återbeskogade områden år 2016. I den sistnämnda skapades en skogsmask med en satellitbild från Landsat 8 och som sedan användes i analysen. Resultaten från analyserna visar att ca 15,8 % av skogen stormfälldes, varav 78 % var barrskog och 13 % var lövskog. År 2016 hade ca 25 % av de stormfällda områdena återbeskogats. Noggrannheten på resultaten är generellt höga men skiljer sig trots detta väsentligt från resultaten i studierna som gjorts för Skogsstyrelsen. Anledningen till att resultaten skiljer sig åt kan bero på vilka satellitbilder och program som använts i analyserna, samt felkällor som uppkommit i samband med analyserna i denna studie. / On January 8th, 2005 one of the most devastating storms in Sweden’s history occurred, where hundreds of thousands became powerless and seven people lost their lives. The storm Gudrun hit central Götaland worst and nearly nine years’ professional felling of forests was estimated to have fallen in some areas. Previous studies of the storm were carried out on behalf of the Swedish Forest Agency, where the results show that the proportion of windthrown forest area was 11 % in the worst affected municipality of Ljungby. About 80 % of all damaged forests were spruce, 18 % were pine and 2 % were deciduous.                        The aim of this thesis is to investigate the amount of windthrown forest after the storm Gudrun through analysis of satellite remote sensing data. The proportion of windthrown coniferous and deciduous forest was calculated and the results were compared to the reports written on behalf of the Swedish Forest Agency. Furthermore, the proportion of reforested areas in 2016 was calculated. A change analysis based on satellite data from Landsat 5 from 2004 and 2005 was performed which included a forest mask created by supervised MLC classification. The forest mask was used to exclude uninteresting areas in the analyses. The result was then used for the analysis of the proportion of windthrown coniferous and deciduous forest and for the analysis of reforested areas in 2016. In the latter, a forest mask based on Landsat 8 data was used. The results from the analyses show that about 15.8 % of the forest was windthrown, of which 78 % were coniferous and 13 % were deciduous forest. By 2016, 25% of the windthrown areas had been reforested. The accuracy of the results is generally high, but despite this, it substantially differs from the results of earlier studies. The reason for this could be differences in satellite images and programs and additional error sources in conjunction with the analyses.

{La première partie de cette thèse est dédiée à la modélisation d'images ou de vidéos considérés comme des fonds sur lesquels on s'attache à détecter des anomalies. Notre analyse de la littérature de la détection d'anomalie sur une seule image nous a fait identifier cinq différentes familles d'hypothèses structurelles sur le fond. Nous proposons de nouveaux algorithmes pour les problèmes de détection d'anomalie sur seule image, de détection de petites cibles sur un fond en mouvement, de détection de changements sur des images satellitaires SAR (Synthetic Aperture Radar) et de détection de nuages dans des séquences d'images de satellite optique.Dans une seconde partie, nous étudions deux autres applications de la modélisation de fond. Pour le débruitage vidéo, nous cherchons pour chaque patch de la vidéo, des patchs similaires le long de la séquence vidéo, et fournissons à un réseau de neurones convolutif les pixels centraux de ces patchs. Le modèle de fond est caché dans les poids du réseau de neurones. Cette méthode s'avère être la plus performante des méthodes par réseau de neurones comparées. Nous étudions également la synthèse de texture à partir d'un exemple. Dans ce problème, des échantillons de texture doivent être générés à partir d'un seul exemple servant de référence. Notre étude distingue les familles d'algorithmes en fonction du type de modèle adopté. Dans le cas des méthodes par réseau de neurones, nous proposons une amélioration corrigeant les artefacts de bord.Dans une troisième partie, nous proposons des implémentations temps-réel GPU de l'interpolation B-spline et de plusieurs algorithmes de débruitage d'images et de vidéo: NL-means, BM3D et VBM3D. La rapidité des implémentations proposées permet leur utilisation dans des scénarios temps-réel, et elles sont en cours de transfert vers l'industrie. / The first part of this thesis is dedicated to the modeling of image or video backgrounds, applied to anomaly detection. In the case of anomaly detection on a single image, our analysis leads us to find five different families of structural assumptions on the background. We propose new algorithms for single-image anomaly detection, small target detection on moving background, change detection on satellite SAR (Synthetic Aperture Radar) images and cloud detection on a sequence of satellite optical images.In the second part, we study two further applications of background modeling. To perform video denoising we search, for every video patch, similar patches in the video sequence, and feed their central pixels to a convolutional neural network (CNN). The background model in this case is hidden in the CNN weights. In our experiments, the proposed method is the best performing of the compared CNN-based methods. We also study exemplar-based texture synthesis. In this problem texture samples have to be generated based on only one reference sample. Our survey classifies the families of algorithms for this task according to their model assumptions. In addition, we propose improvements to fix the border behavior issues that we pointed out in several deep learning based methods.In the third part, we propose real-time GPU implementations for B-spline interpolation and for several image and video denoising algorithms: NL-means, BM3D and VBM3D. The speed of the proposed implementations enables their use in real-time scenarios, and they are currently being transitioned to industry.

abstract: The marked increase in the inflow of remotely sensed data from satellites have trans- formed the Earth and Space Sciences to a data rich domain creating a rich repository for domain experts to analyze. These observations shed light on a diverse array of disciplines ranging from monitoring Earth system components to planetary explo- ration by highlighting the expected trend and patterns in the data. However, the complexity of these patterns from local to global scales, coupled with the volume of this ever-growing repository necessitates advanced techniques to sequentially process the datasets to determine the underlying trends. Such techniques essentially model the observations to learn characteristic parameters of data-generating processes and highlight anomalous planetary surface observations to help domain scientists for making informed decisions. The primary challenge in defining such models arises due to the spatio-temporal variability of these processes. This dissertation introduces models of multispectral satellite observations that sequentially learn the expected trend from the data by extracting salient features of planetary surface observations. The main objectives are to learn the temporal variability for modeling dynamic processes and to build representations of features of interest that is learned over the lifespan of an instrument. The estimated model parameters are then exploited in detecting anomalies due to changes in land surface reflectance as well as novelties in planetary surface landforms. A model switching approach is proposed that allows the selection of the best matched representation given the observations that is designed to account for rate of time-variability in land surface. The estimated parameters are exploited to design a change detector, analyze the separability of change events, and form an expert-guided representation of planetary landforms for prioritizing the retrieval of scientifically relevant observations with both onboard and post-downlink applications. / Dissertation/Thesis / Doctoral Dissertation Computer Engineering 2019

This thesis deals with a development of application running on mobile devices with Windows Mobile operating system. The main task of this application is observing canvas with running presentation and saving time marks, that inform about slide change or animation. Description of system requirements, system analysis using UML language, solutions on image processing level, decription of implementation in C++ language and application tests results are described in the text. Problems of mobile device software development are also outlined in the document. A great part is dedicated to work with multimedia on Windows Mobile 5.0 system, especially to problems linked with DirectShow technology.

No description available.