The natural phenomena simulation in computer graphics is commonly achieved by the procedural methods or the physics model. However, these approaches are hard to directly approach the visual experience. On the other hand, the image reconstruction works can provide the outcome based on real images but lack of interactivity and efficiency on using image resource. For solving these drawbacks, we propose a novel method that enhances the fire simulation effect using the visual learning of image features and generates continuous animations by integrating with procedural methods. We first obtain the dynamics of fire contour by binarization and edge detection. The information extracted from images is gathered into a set of feature data called fire profile. To generate a long sequence of fire animation from a short clip of fire video, we propose two approaches of visual learning to utilize fire profile to produce continuous animation. One is to use the fire image to setup a color value lookup table which contains the average color value of the fire spatial divisions; the other is to design a state machine for describing fire wiggling movement that can generate effects based on user¡¦s input. During the rendering stage of 3D visualization, we set up the fire volume which connecting the feature points of two cross-views by the cubic spline. Then the edge points found on the fire volume can be used as the contour points of the supplementing slices and generate these supplements inside the planned fire volume to formulate a complete fire effect. The proposed method can raise not only the visual reality but also the interactive ability compared with the existing work.

Encontrar las correspondencias entre dos imágenes es un problema crucial en el campo de la visión por ordenador i el reconocimiento de patrones. Es relevante para un amplio rango de propósitos des de aplicaciones de reconocimiento de objetos en las áreas de biometría, análisis de documentos i análisis de formas hasta aplicaciones relacionadas con la geometría desde múltiples puntos de vista tales cómo la recuperación de la pose, estructura desde el movimiento y localización y mapeo. La mayoría de las técnicas existentes enfocan este problema o bien usando características locales en la imagen o bien usando métodos de registro de conjuntos de puntos (o bien una mezcla de ambos). En las primeras, un conjunto disperso de características es primeramente extraído de las imágenes y luego caracterizado en la forma de vectores descriptores usando evidencias locales de la imagen. Las características son asociadas según la similitud entre sus descriptores. En las segundas, los conjuntos de características son considerados cómo conjuntos de puntos los cuales son asociados usando técnicas de optimización no lineal. Estos son procedimientos iterativos que estiman los parámetros de correspondencia y de alineamiento en pasos alternados. Los grafos son representaciones que contemplan relaciones binarias entre las características. Tener en cuenta relaciones binarias al problema de la correspondencia a menudo lleva al llamado problema del emparejamiento de grafos. Existe cierta cantidad de métodos en la literatura destinados a encontrar soluciones aproximadas a diferentes instancias del problema de emparejamiento de grafos, que en la mayoría de casos es del tipo "NP-hard". El cuerpo de trabajo principal de esta tesis está dedicado a formular ambos problemas de asociación de características de imagen y registro de conjunto de puntos como instancias del problema de emparejamiento de grafos. En todos los casos proponemos algoritmos aproximados para solucionar estos problemas y nos comparamos con un número de métodos existentes pertenecientes a diferentes áreas como eliminadores de "outliers", métodos de registro de conjuntos de puntos y otros métodos de emparejamiento de grafos. Los experimentos muestran que en la mayoría de casos los métodos propuestos superan al resto. En ocasiones los métodos propuestos o bien comparten el mejor rendimiento con algún método competidor o bien obtienen resultados ligeramente peores. En estos casos, los métodos propuestos normalmente presentan tiempos computacionales inferiores. / Trobar les correspondències entre dues imatges és un problema crucial en el camp de la visió per ordinador i el reconeixement de patrons. És rellevant per un ampli ventall de propòsits des d’aplicacions de reconeixement d’objectes en les àrees de biometria, anàlisi de documents i anàlisi de formes fins aplicacions relacionades amb geometria des de múltiples punts de vista tals com recuperació de pose, estructura des del moviment i localització i mapeig. La majoria de les tècniques existents enfoquen aquest problema o bé usant característiques locals a la imatge o bé usant mètodes de registre de conjunts de punts (o bé una mescla d’ambdós). En les primeres, un conjunt dispers de característiques és primerament extret de les imatges i després caracteritzat en la forma de vectors descriptors usant evidències locals de la imatge. Les característiques son associades segons la similitud entre els seus descriptors. En les segones, els conjunts de característiques son considerats com conjunts de punts els quals son associats usant tècniques d’optimització no lineal. Aquests son procediments iteratius que estimen els paràmetres de correspondència i d’alineament en passos alternats. Els grafs son representacions que contemplen relacions binaries entre les característiques. Tenir en compte relacions binàries al problema de la correspondència sovint porta a l’anomenat problema de l’emparellament de grafs. Existeix certa quantitat de mètodes a la literatura destinats a trobar solucions aproximades a diferents instàncies del problema d’emparellament de grafs, el qual en la majoria de casos és del tipus “NP-hard”. Una part del nostre treball està dedicat a investigar els beneficis de les mesures de ``bins'' creuats per a la comparació de característiques locals de les imatges. La resta està dedicat a formular ambdós problemes d’associació de característiques d’imatge i registre de conjunt de punts com a instàncies del problema d’emparellament de grafs. En tots els casos proposem algoritmes aproximats per solucionar aquests problemes i ens comparem amb un nombre de mètodes existents pertanyents a diferents àrees com eliminadors d’“outliers”, mètodes de registre de conjunts de punts i altres mètodes d’emparellament de grafs. Els experiments mostren que en la majoria de casos els mètodes proposats superen a la resta. En ocasions els mètodes proposats o bé comparteixen el millor rendiment amb algun mètode competidor o bé obtenen resultats lleugerament pitjors. En aquests casos, els mètodes proposats normalment presenten temps computacionals inferiors.

Šiame darbe pateikiamas apibendrintas automatinio vaizdų jungimo į panoramas algoritmas, detaliai aptariami algoritmo žingsniai ir galimas našumo didinimas lygiagretaus programavimo priemonėmis. Tyrime aprašyti vaizdų jungimo bandymai panaudojant kelias populiarų vaizdų bruožų alternatyvas. Taip pat pristatoma praktinė vaizdų jungimą panaudojanti aplikacija. / This study presents a generalization of an automated image stitching algorithm, describes it's steps in great detail and discusses possible performance improvements by use of parallel execution. Results of stitching experiments using a couple of modern and popular image feature alternatives are shown. A practical application using an implementation of automated image stitching is presented.

This work aims at content-based image search. Different approaches to this type of search are investigated. The main focus of the thesis is special category of content-based image search called sketch-based image search. The most important descriptor types used for image feature extraction in image search are analyzed. Main contribution of the thesis to this research area is a new feature extraction method based on sketch-based image search. This method is implemented together with search interface. The method was evaluated by three test persons. The testing results show promising properties of new method and suggest further possible improve-ments. Powered by TCPDF (www.tcpdf.org)

Information retrieval systems are playing an important role in our day to day life for getting the required information. Many text retrieval systems are available and are working successfully. Even though internet is full of other media like images, audio and video, retrieval systems for these media are rare and have not achieved success as that of text retrieval systems. Image retrieval systems are useful in many applications; there is a high demand for effective and efficient tool for image organization and retrieval as per users need. Images are classified into text based image retrieval and content based image retrieval, we proposed a text based image retrieval system, which makes use of ontology to make the retrieval process intelligent. We worked on Cricket World Cup 2007. We combined text based image retrieval approach with content based image retrieval, which uses color and texture as basic low level features. / kvishu223@gmail.com, pravs72@yahoo.co.in.

Lung cancer is the leading cause of cancer-related death in the United States and worldwide. Early detection of lung cancer can help improve patient outcomes, and survival prediction can inform plans of treatment. By extracting quantitative features from computed tomography scans of lung cancer, predictive models can be built that can achieve both early detection and survival prediction. To build these predictive models, first a detected lung nodule is segmented, then image features are extracted, and finally a model can be built utilizing image features to make predictions. These predictions can help radiologists improve cancer care. Building predictive models based on medical images is the basis of the budding field of radiomics. The hypothesis is that images contain phenotypic information that can be extracted to aid prediction and that automated methods can detect some things beyond human detection. With improved detection and predictive models radiomics aims to help assist radiologists and oncologists provide personalized care. In this work a model is presented to predict long term survival versus short term survival. Forty adenocarcinoma diagnostic lung computed tomography (CT) scans from Moffitt Cancer Center were analyzed for survival prediction. These forty cases were in the top and bottom quartile for survival. A decision tree classifier was able to predict the survival group with an accuracy of 77.5% using five image features chosen from 219 using relief-f. Another contribution of this work is a model for predicting cancer from suspicious nodules. The national lung screening trial was used to build a training set of 261 screening CTs and a test set of 237 CTs. These images were taken at the initial screening, one and two years before cancer developed. From these precursor images, which nodules developed into cancer, could be predicted at 76.79% accuracy with an area under the receiver operating characteristic curve of 0.82. A risk score was also developed to provide a measure of risk during screening. The developed risk score performed favorably in predictive accuracy compared to Lung-RADS on this data set. The Data Science Bowl was also entered and this work examines the knowledge gained from a large-scale competition to improve imaging. In this competition participants were tasked with predicting cancer from 1397 training cases on 506 test cases. The winning entry performed with a logLoss of 0.39975 while making use of all the training data while our entry scored 1.56555 with a different set of training data. A lower logLoss shows greater accuracy. This work explains our approach and examines the winning entry. An overview of the state of radiomicis as it applies to lung cancer is also provided. These contributions of predictive models will help to provide decision support to medical practitioners. By providing tools to the medical field the goal is to advance automated medical imaging to aid clinicians in creating diagnosis and treatment plans.

This thesis is about methods of image comparison. Goal is to create a mobile app that allows user to compare images in real time. In the first part there is a theoretical basis, in particular image similarity algorithms. The practical part contains the app implementation including use case analysis, user interface design and functional requirements. It is followed by source code samples a description of frameworks used. Last part is made of testing the implemented algorithms regarding speed and precision.

The creation, storage, manipulation, and transmission of images have become less costly and more efficient. Consequently, the numbers of images and their users are growing rapidly. This poses challenges to those who organize and provide access to them. One of these challenges is similarity matching. Most current content-based image retrieval (CBIR) systems which can extract only low-level visual features such as color, shape, and texture, use similarity measures based on geometric models of similarity. However, most human similarity judgment data violate the metric axioms of these models. Tversky's (1977) contrast model, which defines similarity as a feature contrast task and equates the degree of similarity of two stimuli to a linear combination of their common and distinctive features, explains human similarity judgments much better than the geometric models. This study tested the contrast model as a conceptual framework to investigate the nature of the relationships between features and similarity of images as perceived by human judges. Data were collected from 150 participants who performed two tasks: an image description and a similarity judgment task. Qualitative methods (content analysis) and quantitative (correlational) methods were used to seek answers to four research questions related to the relationships between common and distinctive features and similarity judgments of images as well as measures of their common and distinctive features. Structural equation modeling, correlation analysis, and regression analysis confirmed the relationships between perceived features and similarity of objects hypothesized by Tversky (1977). Tversky's (1977) contrast model based upon a combination of two methods for measuring common and distinctive features, and two methods for measuring similarity produced statistically significant structural coefficients between the independent latent variables (common and distinctive features) and the dependent latent variable (similarity). This model fit the data well for a sample of 30 (435 pairs of) images and 150 participants (χ2 =16.97, df=10, p = .07508, RMSEA= .040, SRMR= .0205, GFI= .990, AGFI= .965). The goodness of fit indices showed the model did not significantly deviate from the actual sample data. This study is the first to test the contrast model in the context of information representation and retrieval. Results of the study are hoped to provide the foundations for future research that will attempt to further test the contrast model and assist designers of image organization and retrieval systems by pointing toward alternative document representations and similarity measures that more closely match human similarity judgments.

An AdaBoost algorithm for construction of strong classifier from several weak hypotesis will be presented in this work. Theoretical background of the algorithm and the method of construction of strong classifiers will be explained. WaldBoost extension to the algorithm will be described. The thesis deals with image features that are often used as element of weak classifiers. Brief introduction to pattern recognition in context of computer vision will be outlined in the begining of the work. Also some widely used methods of classifier training will be presented. An object detection library based on AdaBoost classifiers was developed as part of the work. The library was used in implementation of software that in praktice demonstrates object detection in videosquences. Last part of the work describes tool for training of AdaBoost classifiers.

Computational creativity is a new and developing field of artificial intelligence concerned with computational systems that either autonomously produce original and functional products, or that augment the ability of humans to do so. As the role of computers in our daily lives is continuing to expand, the need for such systems is becoming increasingly important. We introduce and document the development of a new “creative” system, called DARCI (Digital ARtist Communicating Intention), that is designed to autonomously create novel artistic images that convey linguistic concepts to the viewer. Within the scope of this work, the system becomes capable of creating non-photorealistic renderings of existing image compositions so that they convey the semantics of given adjectives. Ultimately, we show that DARCI is capable of producing surprising artifacts that are competitive, in some ways, with those produced by human artists. As with the development of any “creative” system, we are faced with the challenges of incorporating the philosophies of creativity into the design of the system, assessing the system's creativity, overcoming technical shortcomings of extant modern algorithms, and justifying the system within its creative domain (in this case, visual art). In meeting these challenges with DARCI, we demonstrate three broad contributions of the system: 1) the contribution to the field of computational creativity in the form of an original system, new approaches to achieving autonomy in creative systems, and new practical assessment methods; 2) the contribution to the field of computer vision in the form of new image features for affective image annotation and a new dataset; and 3) the contribution to the domain of visual art in the form of mutually beneficial collaborations and participation in several art galleries and exhibits.