Image analysis is one of the most challenging areas in today's computer sci ence, and image technologies are used in a host of applications. This book concentrates on image textures and presents novel techniques for their sim ulation, retrieval, and segmentation using specific Gibbs random fields with multiple pairwise interaction between signals as probabilistic image models. These models and techniques were developed mainly during the previous five years (in relation to April 1999 when these words were written). While scanning these pages you may notice that, in spite of long equa tions, the mathematical background is extremely simple. I have tried to avoid complex abstract constructions and give explicit physical (to be spe cific, "image-based") explanations to all the mathematical notions involved. Therefore it is hoped that the book can be easily read both by professionals and graduate students in computer science and electrical engineering who take an interest in image analysis and synthesis. Perhaps, mathematicians studying applications of random fields may find here some less traditional, and thus controversial, views and techniques.

This text presents techniques for describing image textures. Contrary to the usual practice of embedding the images to known modelling frameworks borrowed from statistical physics or other domains, this book deduces the Gibbs models from basic image features and tailors the modelling framework to the images. This approach results in more general Gibbs models than can be either Markovian or non-Markovian and possess arbitrary interaction structures and strengths. The book presents computationally feasible algorithms for parameter estimation and image simulation and demonstrates their abilities and limitations by numerous experimental results.

"This book is concerned with a probabilistic approach for image analysis, mostly from the Bayesian point of view, and the important Markov chain Monte Carlo methods commonly used....This book will be useful, especially to researchers with a strong background in probability and an interest in image analysis. The author has presented the theory with rigor...he doesn’t neglect applications, providing numerous examples of applications to illustrate the theory." -- MATHEMATICAL REVIEWS

Markov random field (MRF) theory provides a basis for modeling contextual constraints in visual processing and interpretation. It enables us to develop optimal vision algorithms systematically when used with optimization principles. This book presents a comprehensive study on the use of MRFs for solving computer vision problems. The book covers the following parts essential to the subject: introduction to fundamental theories, formulations of MRF vision models, MRF parameter estimation, and optimization algorithms. Various vision models are presented in a unified framework, including image restoration and reconstruction, edge and region segmentation, texture, stereo and motion, object matching and recognition, and pose estimation. This second edition includes the most important progress in Markov modeling in image analysis in recent years such as Markov modeling of images with "macro" patterns (e.g. the FRAME model), Markov chain Monte Carlo (MCMC) methods, reversible jump MCMC. This book is an excellent reference for researchers working in computer vision, image processing, statistical pattern recognition and applications of MRFs. It is also suitable as a text for advanced courses in these areas.

This volume contains the ?nal version of the papers originally presented at the second SMILE workshop 3D Structure from Multiple Images of Large-scale Environments, which was held on 1-2 July 2000 in conjunction with the Sixth European Conference in Computer Vision at Trinity College Dublin. The subject of the workshop was the visual acquisition of models of the 3D world from images and their application to virtual and augmented reality. Over the last few years tremendous progress has been made in this area. On the one hand important new insightshavebeenobtainedresultinginmore exibilityandnewrepresentations.Onthe other hand a number of techniques have come to maturity, yielding robust algorithms delivering good results on real image data. Moreover supporting technologies – such as digital cameras, computers, disk storage, and visualization devices – have made things possible that were infeasible just a few years ago. Opening the workshop was Paul Debevec s invited presentation on image-based modeling,rendering,andlighting.Hepresentedanumberoftechniquesforusingdigital images of real scenes to create 3D models, virtual camera moves, and realistic computer animations.Theremainderoftheworkshopwasdividedintothreesessions:Computation and Algorithms, Visual Scene Representations, and Extended Environments. After each session there was a panel discussion that included all speakers. These panel discussions were organized by Bill Triggs, Marc Pollefeys, and Tomas Pajdla respectively, who introduced the topics and moderated the discussion. Asubstantialpartoftheseproceedingsarethetranscriptsofthediscussionsfollowing each paper and the full panel sessions. These discussions were of very high quality and were an integral part of the workshop.

Texture analysis is one of the fundamental aspects of human vision by which we discriminate between surfaces and objects. In a similar manner, computer vision can take advantage of the cues provided by surface texture to distinguish and recognize objects. In computer vision, texture analysis may be used alone or in combination with other sensed features (e.g. color, shape, or motion) to perform the task of recognition. Either way, it is a feature of paramount importance and boasts a tremendous body of work in terms of both research and applications.Currently, the main approaches to texture analysis must be sought out through a variety of research papers. This collection of chapters brings together in one handy volume the major topics of importance, and categorizes the various techniques into comprehensible concepts. The methods covered will not only be relevant to those working in computer vision, but will also be of benefit to the computer graphics, psychophysics, and pattern recognition communities, academic or industrial.

Imaging & Vision Systems - Theory, Assessment & Applications, Advances in Computation, Theory & Practice -- Volume 9

This useful textbook/reference presents an accessible primer on the fundamentals of image texture analysis, as well as an introduction to the K-views model for extracting and classifying image textures. Divided into three parts, the book opens with a review of existing models and algorithms for image texture analysis, before delving into the details of the K-views model. The work then concludes with a discussion of popular deep learning methods for image texture analysis. Topics and features: provides self-test exercises in every chapter; describes the basics of image texture, texture features, and image texture classification and segmentation; examines a selection of widely-used methods for measuring and extracting texture features, and various algorithms for texture classification; explains the concepts of dimensionality reduction and sparse representation; discusses view-based approaches to classifying images; introduces the template for the K-views algorithm, as well as a range of variants of this algorithm; reviews several neural network models for deep machine learning, and presents a specific focus on convolutional neural networks. This introductory text on image texture analysis is ideally suitable for senior undergraduate and first-year graduate students of computer science, who will benefit from the numerous clarifying examples provided throughout the work.

This text is concerned with a probabilistic approach to image analysis as initiated by U. GRENANDER, D. and S. GEMAN, B.R. HUNT and many others, and developed and popularized by D. and S. GEMAN in a paper from 1984. It formally adopts the Bayesian paradigm and therefore is referred to as 'Bayesian Image Analysis'. There has been considerable and still growing interest in prior models and, in particular, in discrete Markov random field methods. Whereas image analysis is replete with ad hoc techniques, Bayesian image analysis provides a general framework encompassing various problems from imaging. Among those are such 'classical' applications like restoration, edge detection, texture discrimination, motion analysis and tomographic reconstruction. The subject is rapidly developing and in the near future is likely to deal with high-level applications like object recognition. Fascinating experiments by Y. CHOW, U. GRENANDER and D.M. KEENAN (1987), (1990) strongly support this belief.