Billedbehandling

Gaussian scale-space is one of the best understood multi-resolution techniques available to the computer vision and image analysis community. It is the purpose of this book to guide the reader through some of its main aspects. During an intensive weekend in May 1996 a workshop on Gaussian scale-space theory was held in Copenhagen, which was attended by many of the leading experts in the field. The bulk of this book originates from this workshop. Presently there exist only two books on the subject. In contrast to Lindeberg's monograph (Lindeberg, 1994e) this book collects contributions from several scale- space researchers, whereas it complements the book edited by ter Haar Romeny (Haar Romeny, 1994) on non-linear techniques by focusing on linear diffusion. This book is divided into four parts. The reader not so familiar with scale-space will find it instructive to first consider some potential applications described in Part 1. Parts II and III both address fundamental aspects of scale-space. Whereas scale is treated as an essentially arbitrary constant in the former, the latter em- phasizes the deep structure, i.e. the structure that is revealed by varying scale. Finally, Part IV is devoted to non-linear extensions, notably non-linear diffusion techniques and morphological scale-spaces, and their relation to the linear case. The Danish National Science Research Council is gratefully acknowledged for providing financial support for the workshop under grant no. 9502164.

The past decades have seen significant improvements in 3D imaging where the related techniques and technologies have advanced to a mature state. These exciting developments have sparked increasing interest in industry and academia in the challenges and opportunities afforded by 3D sensing. As a consequence, the emerging area of safety and security related imaging incorporates these important new technologies beyond the limitations of 2D image processing. This book is so far the first that covers the current state of the art in 3D imaging for safety and security. It reports about selected contributions given at the "e;Workshop on Advanced 3D Imaging for Safety and Security"e; held in conjunction with the International Conference on Computer Vision and Pattern Recognition CVPR 2005, June 2005, San Diego, CA. The workshop brought together pioneering academic and industrial researchers in the field of computer vision and image analysis. Special attention was given to advanced 3D imaging technologies in the context of safety and security applications. Comparative evaluation studies showing advantages of 3D imaging over traditional 2D imaging for a given computer vision or pattern recognition task were emphasized. Moreover, additional experts in the field of 3D imaging for safety and security were invited by the editors for a contribution to this book. The book is structured into two parts, each containing five or six chapters on (1) Biometrics and (2) Safety and Security Applications.

From grading and preparing harvested vegetables to the tactile probing of a patient's innermost recesses, mechatronics has become part of our way of life. The addition of senses and computing intelligence to blend with mechanical actuation gives rise to a breed of new machines with all the best attributes of a robot. Here we find educational robots competing under water and dancing on land. Surgical robots drill precision holes in the skull while others direct radiation treatment or extract blood from an umbilical cord. Machine vision manages beer kegs and automated paint spraying, while controlling feral animals and the watering of cotton crops. Although there is no shortage of theoretical and technical detail in these chapters, they have the common theme that they describe work that has been applied in practice. They are vital reading both for students of mechatronics and for engineers harnessing its power to create new products.

The two-volume set of LNCS 10941 and 10942 constitutes the proceedings of the 9th International Conference on Advances in Swarm Intelligence, ICSI 2018, held in Shanghai, China, in June 2018. The total of 113 papers presented in these volumes was carefully reviewed and selected from 197 submissions. The papers were organized in topical sections namely: multi-agent systems; swarm robotics; fuzzy logic approaches; planning and routing problems; recommendation in social media; predication; classification; finding patterns; image enhancement; deep learning; theories and models of swarm intelligence; ant colony optimization; particle swarm optimization; artificial bee colony algorithms; genetic algorithms; differential evolution; fireworks algorithm; bacterial foraging optimization; artificial immune system; hydrologic cycle optimization; other swarm-based optimization algorithms; hybrid optimization algorithms; multi-objective optimization; large-scale global optimization.

As we increase our reliance on computer-generated information, often using it as part of our decision-making process, we must devise tools to assess the correctness of that information. Consider, for example, software embedded on vehicles, used for simulating aircraft performance, or used in medical imaging. In those cases, software correctness is of paramount importance as there's little room for error. Software verification is one of the tools available to attain such goals. Verification is a well known and widely studied subfield of computer science and computational science and the goal is to help us increase confidence in the software implementation by verifying that the software does what it is supposed to do. The goal of this book is to introduce the reader to software verification in the context of visualization. In the same way we became more dependent on commercial software, we have also increased our reliance on visualization software. The reason is simple: visualization isthe lens through which users can understand complex data, and as such it must be verified. The explosion in our ability to amass data requires tools not only to store and analyze data, but also to visualize it. This book is comprised of six chapters. After an introduction to the goals of the book, we present a brief description of both worlds of visualization (Chapter 2) and verification (Chapter 3). We then proceed to illustrate the main steps of the verification pipeline for visualization algorithms. We focus on two classic volume visualization techniques, namely, Isosurface Extraction (Chapter 4) and Direct Volume Rendering (Chapter 5). We explain how to verify implementations of those techniques and report the latest results in the field of verification of visualization techniques. The last chapter concludes the book and highlights new research topics for the future.

Path planning and navigation are indispensable components for controlling autonomous agents in interactive virtual worlds. Given the growing demands on the size and complexity of modern virtual worlds, a number of new techniques have been developed for achieving intelligent navigation for the next generation of interactive multi-agent simulations. This book reviews the evolution of several related techniques, starting from classical planning and computational geometry techniques and then gradually moving toward more advanced topics with focus on recent developments from the work of the authors. The covered topics range from discrete search and geometric representations to planning under different types of constraints and harnessing the power of graphics hardware in order to address Euclidean shortest paths and discrete search for multiple agents under limited time budgets. The use of planning algorithms beyond path planning is also discussed in the areas of crowd animation and whole-body motion planning for virtual characters.

In image processing and computer vision applications such as medical or scientific image data analysis, as well as in industrial scenarios, images are used as input measurement data. It is good scientific practice that proper measurements must be equipped with error and uncertainty estimates. For many applications, not only the measured values but also their errors and uncertainties, should be-and more and more frequently are-taken into account for further processing. This error and uncertainty propagation must be done for every processing step such that the final result comes with a reliable precision estimate. The goal of this book is to introduce the reader to the recent advances from the field of uncertainty quantification and error propagation for computer vision, image processing, and image analysis that are based on partial differential equations (PDEs). It presents a concept with which error propagation and sensitivity analysis can be formulated with a set of basic operations. The approach discussed in this book has the potential for application in all areas of quantitative computer vision, image processing, and image analysis. In particular, it might help medical imaging finally become a scientific discipline that is characterized by the classical paradigms of observation, measurement, and error awareness. This book is comprised of eight chapters. After an introduction to the goals of the book (Chapter 1), we present a brief review of PDEs and their numerical treatment (Chapter 2), PDE-based image processing (Chapter 3), and the numerics of stochastic PDEs (Chapter 4). We then proceed to define the concept of stochastic images (Chapter 5), describe how to accomplish image processing and computer vision with stochastic images (Chapter 6), and demonstrate the use of these principles for accomplishing sensitivity analysis (Chapter 7). Chapter 8 concludes the book and highlights new research topics for the future.

New data acquisition techniques are emerging and are providing fast and efficient means for multidimensional spatial data collection. Airborne LIDAR surveys, SAR satellites, stereo-photogrammetry and mobile mapping systems are increasingly used for the digital reconstruction of the environment. All these systems provide extremely high volumes of raw data, often enriched with other sensor data (e.g., beam intensity). Improving methods to process and visually analyze this massive amount of geospatial and user-generated data is crucial to increase the efficiency of organizations and to better manage societal challenges. Within this context, this book proposes an up-to-date view of computational methods and tools for spatio-temporal data fusion, multivariate surface generation, and feature extraction, along with their main applications for surface approximation and rainfall analysis. The book is intended to attract interest from different fields, such as computer vision, computer graphics, geomatics, and remote sensing, working on the common goal of processing 3D data. To this end, it presents and compares methods that process and analyze the massive amount of geospatial data in order to support better management of societal challenges through more timely and better decision making, independent of a specific data modeling paradigm (e.g., 2D vector data, regular grids or 3D point clouds). We also show how current research is developing from the traditional layered approach, adopted by most GIS softwares, to intelligent methods for integrating existing data sets that might contain important information on a geographical area and environmental phenomenon. These services combine traditional map-oriented visualization with fully 3D visual decision support methods and exploit semantics-oriented information (e.g., a-priori knowledge, annotations, segmentations) when processing, merging, and integrating big pre-existing data sets.

Physics-based animation is commonplace in animated feature films and even special effects for live-action movies. Think about a recent movie and there will be some sort of special effects such as explosions or virtual worlds. Cloth simulation is no different and is ubiquitous because most virtual characters (hopefully!) wear some sort of clothing.The focus of this book is physics-based cloth simulation. We start by providing background information and discuss a range of applications. This book provides explanations of multiple cloth simulation techniques. More specifically, we start with the most simple explicitly integrated mass-spring model and gradually work our way up to more complex and commonly used implicitly integrated continuum techniques in state-of-the-art implementations. We give an intuitive explanation of the techniques and give additional information on how to efficiently implement them on a computer.This book discusses explicit and implicit integration schemes for cloth simulation modeled with mass-spring systems. In addition to this simple model, we explain the more advanced continuum-inspired cloth model introduced in the seminal work of Baraff and Witkin [1998]. This method is commonly used in industry.We also explain recent work by Liu et al. [2013] that provides a technique to obtain fast simulations. In addition to these simulation approaches, we discuss how cloth simulations can be art directed for stylized animations based on the work of Wojan et al. [2016]. Controllability is an essential component of a feature animation film production pipeline. We conclude by pointing the reader to more advanced techniques.

In geometry processing and shape analysis, several applications have been addressed through the properties of the Laplacian spectral kernels and distances, such as commute time, biharmonic, diffusion, and wave distances.Within this context, this book is intended to provide a common background on the definition and computation of the Laplacian spectral kernels and distances for geometry processing and shape analysis. To this end, we define a unified representation of the isotropic and anisotropic discrete Laplacian operator on surfaces and volumes; then, we introduce the associated differential equations, i.e., the harmonic equation, the Laplacian eigenproblem, and the heat equation. Filtering the Laplacian spectrum, we introduce the Laplacian spectral distances, which generalize the commute-time, biharmonic, diffusion, and wave distances, and their discretization in terms of the Laplacian spectrum. As main applications, we discuss the design of smooth functions and the Laplacian smoothing of noisy scalar functions.All the reviewed numerical schemes are discussed and compared in terms of robustness, approximation accuracy, and computational cost, thus supporting the reader in the selection of the most appropriate with respect to shape representation, computational resources, and target application.

This synthesis lecture presents an intuitive introduction to the mathematics of motion and deformation in computer graphics. Starting with familiar concepts in graphics, such as Euler angles, quaternions, and affine transformations, we illustrate that a mathematical theory behind these concepts enables us to develop the techniques for efficient/effective creation of computer animation.This book, therefore, serves as a good guidepost to mathematics (differential geometry and Lie theory) for students of geometric modeling and animation in computer graphics. Experienced developers and researchers will also benefit from this book, since it gives a comprehensive overview of mathematical approaches that are particularly useful in character modeling, deformation, and animation.

This book serves as a practical guide to simulation of 3D deformable solids using the Finite Element Method (FEM). It reviews a number of topics related to the theory and implementation of FEM approaches: measures of deformation, constitutive laws of nonlinear materials, tetrahedral discretizations, and model reduction techniques for real-time simulation. Simulations of deformable solids are important in many applications in computer graphics, including film special effects, computer games, and virtual surgery. The Finite Element Method has become a popular tool in many such applications. Variants of FEM catering to both offline and real-time simulation have had a mature presence in computer graphics literature. This book is designed for readers familiar with numerical simulation in computer graphics, who would like to obtain a cohesive picture of the various FEM simulation methods available, their strengths and weaknesses, and their applicability in various simulation scenarios. The book is also a practical implementation guide for the visual effects developer, offering a lean yet adequate synopsis of the underlying mathematical theory. Chapter 1 introduces the quantitative descriptions used to capture the deformation of elastic solids, the concept of strain energy, and discusses how force and stress result as a response to deformation. Chapter 2 reviews a number of constitutive models, i.e., analytical laws linking deformation to the resulting force that has successfully been used in various graphics-oriented simulation tasks. Chapter 3 summarizes how deformation and force can be computed discretely on a tetrahedral mesh, and how an implicit integrator can be structured around this discretization. Finally, chapter 4 presents the state of the art in model reduction techniques for real-time FEM solid simulation and discusses which techniques are suitable for which applications. Topics discussed in this chapter include linear modal analysis, modal warping, subspace simulation, and domain decomposition.

Heritage sites across the world have witnessed a number of natural calamities, sabotage and damage from visitors, resulting in their present ruined condition. Many sites are now restricted to reduce the risk of further damage. Yet these masterpieces are significant cultural icons and critical markers of past civilizations that future generations need to see. A digitally reconstructed heritage site could diminish further harm by using immersive navigation or walkthrough systems for virtual environments. An exciting key element for the viewer is observing fine details of the historic work and viewing monuments in their undamaged form. This book presents image super-resolution methods and techniques for automatically detecting and inpainting damaged regions in heritage monuments, in order to provide an enhanced visual experience.The book presents techniques to obtain higher resolution photographs of the digitally reconstructed monuments, and the resulting images can serve as input to immersive walkthrough systems. It begins with the discussion of two novel techniques for image super-resolution and an approach for inpainting a user-supplied region in the given image, followed by a technique to simultaneously perform super-resolution and inpainting of given missing regions. It then introduces a method for automatically detecting and repairing the damage to dominant facial regions in statues, followed by a few approaches for automatic crack repair in images of heritage scenes. This book is a giant step toward ensuring that the iconic sites of our past are always available, and will never be truly lost.

This book provides beginners in computer graphics and related fields a guide to the concepts, models, and technologies for realistic rendering of material appearance. It provides a complete and thorough overview of reflectance models and acquisition setups, along with providing a selection of the available tools to explore, visualize, and render the reflectance data. Reflectance models are under continuous development, since there is still no straightforward solution for general material representations. Every reflectance model is specific to a class of materials. Hence, each has strengths and weaknesses, which the book highlights in order to help the reader choose the most suitable model for any purpose. The overview of the acquisition setups will provide guidance to a reader who needs to acquire virtual materials and will help them to understand which measurement setup can be useful for a particular purpose, while taking into account the performance and the expected cost derived from the required components. The book also describes several recent open source software solutions, useful for visualizing and manipulating a wide variety of reflectance models and data.

As new displays and cameras offer enhanced color capabilities, there is a need to extend the precision of digital content. High Dynamic Range (HDR) imaging encodes images and video with higher than normal 8 bit-per-color-channel precision, enabling representation of the complete color gamut and the full visible range of luminance.However, to realize transition from the traditional toHDRimaging, it is necessary to develop imaging algorithms that work with the high-precision data. Tomake such algorithms effective and feasible in practice, it is necessary to take advantage of the limitations of the human visual system by aligning the data shortcomings to those of the human eye, thus limiting storage and processing precision. Therefore, human visual perception is the key component of the solutions we discuss in this book. This book presents a complete pipeline forHDR image and video processing fromacquisition, through compression and quality evaluation, to display. At the HDR image and video acquisition stage specialized HDR sensors or multi-exposure techniques suitable for traditional cameras are discussed. Then, we present a practical solution for pixel values calibration in terms of photometric or radiometric quantities, which are required in some technically oriented applications. Also, we cover the problem of efficient image and video compression and encoding either for storage or transmission purposes, including the aspect of backward compatibility with existing formats. Finally, we review existing HDR display technologies and the associated problems of image contrast and brightness adjustment. For this purpose tone mapping is employed to accommodate HDR content to LDR devices. Conversely, the so-called inverse tone mapping is required to upgrade LDR content for displaying on HDR devices. We overview HDR-enabled image and video quality metrics, which are needed to verify algorithms at all stages of the pipeline. Additionally, we cover successful examples of the HDR technology applications, in particular, in computer graphics and computer vision. The goal of this book is to present all discussed components of the HDR pipeline with the main focus on video. For some pipeline stages HDR video solutions are either not well established or do not exist at all, in which case we describe techniques for single HDR images. In such cases we attempt to select the techniques, which can be extended into temporal domain. Whenever needed, relevant background information on human perception is given, which enables better understanding of the design choices behind the discussed algorithms and HDR equipment. Table of Contents: Introduction / Representation of an HDR Image / HDR Image and Video Acquisition / HDR Image Quality / HDR Image, Video, and Texture Compression / Tone Reproduction / HDR Display Devices / LDR2HDR: Recovering Dynamic Range in Legacy Content / HDRI in Computer Graphics / Software

Information theory (IT) tools, widely used in scientific fields such as engineering, physics, genetics, neuroscience, and many others, are also emerging as useful transversal tools in computer graphics. In this book, we present the basic concepts of IT and how they have been applied to the graphics areas of radiosity, adaptive ray-tracing, shape descriptors, viewpoint selection and saliency, scientific visualization, and geometry simplification. Some of the approaches presented, such as the viewpoint techniques, are now the state of the art in visualization. Almost all of the techniques presented in this book have been previously published in peer-reviewed conference proceedings or international journals. Here, we have stressed their common aspects and presented them in an unified way, so the reader can clearly see which problems IT tools can help solve, which specific tools to use, and how to apply them. A basic level of knowledge in computer graphics is required but basic concepts in IT are presented. The intended audiences are both students and practitioners of the fields above and related areas in computer graphics. In addition, IT practitioners will learn about these applications. Table of Contents: Information Theory Basics / Scene Complexity and Refinement Criteria for Radiosity / Shape Descriptors / Refinement Criteria for Ray-Tracing / Viewpoint Selection and Mesh Saliency / View Selection in Scientific Visualization / Viewpoint-based Geometry Simplification

Tiling theory is an elegant branch of mathematics that has applications in several areas of computer science. The most immediate application area is graphics, where tiling theory has been used in the contexts of texture generation, sampling theory, remeshing, and of course the generation of decorative patterns. The combination of a solid theoretical base (complete with tantalizing open problems), practical algorithmic techniques, and exciting applications make tiling theory a worthwhile area of study for practitioners and students in computer science. This synthesis lecture introduces the mathematical and algorithmic foundations of tiling theory to a computer graphics audience. The goal is primarily to introduce concepts and terminology, clear up common misconceptions, and state and apply important results. The book also describes some of the algorithms and data structures that allow several aspects of tiling theory to be used in practice. Table of Contents: Introduction / Tiling Basics / Symmetry / Tilings by Polygons / Isohedral Tilings / Nonperiodic and Aperiodic Tilings / Survey

Computer vision has become increasingly important and effective in recent years due to its wide-ranging applications in areas as diverse as smart surveillance and monitoring, health and medicine, sports and recreation, robotics, drones, and self-driving cars. Visual recognition tasks, such as image classification, localization, and detection, are the core building blocks of many of these applications, and recent developments in Convolutional Neural Networks (CNNs) have led to outstanding performance in these state-of-the-art visual recognition tasks and systems. As a result, CNNs now form the crux of deep learning algorithms in computer vision.This self-contained guide will benefit those who seek to both understand the theory behind CNNs and to gain hands-on experience on the application of CNNs in computer vision. It provides a comprehensive introduction to CNNs starting with the essential concepts behind neural networks: training, regularization, and optimization of CNNs.The book also discusses a wide range of loss functions, network layers, and popular CNN architectures, reviews the different techniques for the evaluation of CNNs, and presents some popular CNN tools and libraries that are commonly used in computer vision. Further, this text describes and discusses case studies that are related to the application of CNN in computer vision, including image classification, object detection, semantic segmentation, scene understanding, and image generation.This book is ideal for undergraduate and graduate students, as no prior background knowledge in the field is required to follow the material, as well as new researchers, developers, engineers, and practitioners who are interested in gaining a quick understanding of CNN models.

For the last ten years, face biometric research has been intensively studied by the computer vision community. Face recognition systems have been used in mobile, banking, and surveillance systems. For face recognition systems, face spoofing attack detection is a crucial stage that could cause severe security issues in government sectors. Although effective methods for face presentation attack detection have been proposed so far, the problem is still unsolved due to the difficulty in the design of features and methods that can work for new spoofing attacks. In addition, existing datasets for studying the problem are relatively small which hinders the progress in this relevant domain.In order to attract researchers to this important field and push the boundaries of the state of the art on face anti-spoofing detection, we organized the Face Spoofing Attack Workshop and Competition at CVPR 2019, an event part of the ChaLearn Looking at People Series. As part of this event, we released the largest multi-modal face anti-spoofing dataset so far, the CASIA-SURF benchmark. The workshop reunited many researchers from around the world and the challenge attracted more than 300 teams. Some of the novel methodologies proposed in the context of the challenge achieved state-of-the-art performance. In this manuscript, we provide a comprehensive review on face anti-spoofing techniques presented in this joint event and point out directions for future research on the face anti-spoofing field.

Being able to recover the shape of 3D deformable surfaces from a single video stream would make it possible to field reconstruction systems that run on widely available hardware without requiring specialized devices. However, because many different 3D shapes can have virtually the same projection, such monocular shape recovery is inherently ambiguous. In this survey, we will review the two main classes of techniques that have proved most effective so far: The template-based methods that rely on establishing correspondences with a reference image in which the shape is already known, and non-rigid structure-from-motion techniques that exploit points tracked across the sequences to reconstruct a completely unknown shape. In both cases, we will formalize the approach, discuss its inherent ambiguities, and present the practical solutions that have been proposed to resolve them. To conclude, we will suggest directions for future research. Table of Contents: Introduction / Early Approaches to Non-Rigid Reconstruction / Formalizing Template-Based Reconstruction / Performing Template-Based Reconstruction / Formalizing Non-Rigid Structure from Motion / Performing Non-Rigid Structure from Motion / Future Directions

Face detection, because of its vast array of applications, is one of the most active research areas in computer vision. In this book, we review various approaches to face detection developed in the past decade, with more emphasis on boosting-based learning algorithms. We then present a series of algorithms that are empowered by the statistical view of boosting and the concept of multiple instance learning. We start by describing a boosting learning framework that is capable to handle billions of training examples. It differs from traditional bootstrapping schemes in that no intermediate thresholds need to be set during training, yet the total number of negative examples used for feature selection remains constant and focused (on the poor performing ones). A multiple instance pruning scheme is then adopted to set the intermediate thresholds after boosting learning. This algorithm generates detectors that are both fast and accurate. We then present two multiple instance learning schemes for face detection, multiple instance learning boosting (MILBoost) and winner-take-all multiple category boosting (WTA-McBoost). MILBoost addresses the uncertainty in accurately pinpointing the location of the object being detected, while WTA-McBoost addresses the uncertainty in determining the most appropriate subcategory label for multiview object detection. Both schemes can resolve the ambiguity of the labeling process and reduce outliers during training, which leads to improved detector performances. In many applications, a detector trained with generic data sets may not perform optimally in a new environment. We propose detection adaption, which is a promising solution for this problem. We present an adaptation scheme based on the Taylor expansion of the boosting learning objective function, and we propose to store the second order statistics of the generic training data for future adaptation. We show that with a small amount of labeled data in the new environment, the detector's performance can be greatly improved. We also present two interesting applications where boosting learning was applied successfully. The first application is face verification for filtering and ranking image/video search results on celebrities. We present boosted multi-task learning (MTL), yet another boosting learning algorithm that extends MILBoost with a graphical model. Since the available number of training images for each celebrity may be limited, learning individual classifiers for each person may cause overfitting. MTL jointly learns classifiers for multiple people by sharing a few boosting classifiers in order to avoid overfitting. The second application addresses the need of speaker detection in conference rooms. The goal is to find who is speaking, given a microphone array and a panoramic video of the room. We show that by combining audio and visual features in a boosting framework, we can determine the speaker's position very accurately. Finally, we offer our thoughts on future directions for face detection. Table of Contents: A Brief Survey of the Face Detection Literature / Cascade-based Real-Time Face Detection / Multiple Instance Learning for Face Detection / Detector Adaptation / Other Applications / Conclusions and Future Work