Given a familiar object extracted from its surroundings, we humans have little difficulty in recognizing it irrespective of its size, position and orientation in our field of view. Changes in lighting and the effects of perspective also pose no problems. How do we achieve this, and more importantly, how can we get a computer to do this? One very promising approach is to find mathematical functions of an object's image, or of an object's 3D description, that are invariant to the transformations caused by the object's motion. This book is devoted to the theory and practice of such invariant image features, so-called image invariants, for planar objects. It gives a comprehensive summary of the field, discussing methods for recognizing both occluded and partially occluded objects, and also contains a definitive treatmentof moment invariants and a tutorial introduction to algebraic invariants, which are fundamental to affine moment invariants and to many projective invariants. A number of novel invariant functions are presented and the results of numerous experiments investigating the stability of new and old invariants are discussed. The main conclusion is that moment invariants are very effective, both for partially occluded objects and for recognizing objects in grey-level images.

The papers contained in this volume were presented at the Fourth Annual Symposium on Combinatorial Pattern Matching, held in Padova, Italy, in June 1993. Combinatorial pattern matching addresses issues of searching and matching of strings and more complicated patterns such as trees, regular expressions, extended expressions, etc. The goal is to derive nontrivial combinatorial properties for such structures and then to exploit these properties in order to achieve superior performance for the corresponding computational problems. In recent years, a steady flow of high-quality scientific studies of this subject has changed a sparse set of isolated results into a full-fledged area of algorithmics. The area is expected to grow even further due to the increasing demand for speedand efficiency that comes especially from molecular biology and the Genome project, but also from other diverse areas such as information retrieval, pattern recognition, compilers, data compression, and program analysis.

Multi-Modal User Interactions in Controlled Environments investigates the capture and analysis of user's multimodal behavior (mainly eye gaze, eye fixation, eye blink and body movements) within a real controlled environment (controlled-supermarket, personal environment) in order to adapt the response of the computer/environment to the user. Such data is captured using non-intrusive sensors (for example, cameras in the stands of a supermarket) installed in the environment. This multi-modal video based behavioral data will be analyzed to infer user intentions while assisting users in their day-to-day tasks by adapting the system's response to their requirements seamlessly. This book also focuses on the presentation of information to the user. Multi-Modal User Interactions in Controlled Environments is designed for professionals in industry, including professionals in the domains of security and interactive web television. This book is also suitable for graduate-level students in computer science and electrical engineering.

This volume contains the 22 papers accepted for presentation at the Third Annual Symposium on Combinatorial Pattern Matching held April 29 to May 1, 1992, in Tucson, Arizona; it constitutes the first conference proceedings entirely devoted to combinatorial pattern matching (CPM). CPM deals withissues of searching and matching of strings and other more complicated patterns such as trees, regular expressions, extended expressions, etc. in order to derive combinatorial properties for such structures. As an interdisciplinary field of growing interest, CPM is related to research in information retrieval, pattern recognition, compilers, data compression, and program analysis as well as to results, problems and methods from combinatorial mathematics and molecular biology.

This volume contains the papers selected for presentation at the Second International Conference on Parallel Image Analysis (ICPIA '92), held in Ube, Japan, December 21-23, 1992. The conference topics are data structures, parallel algorithms and architectures, neural networks, computational vision, syntactic generation and recognition, and multidimensional models. The first meeting with these topics was theInternational Colloquium on Parallel Image Processing, which took place in Paris in June 1991. The aim of the meetings is to bring together specialistsfrom various countries who are interested in the topics and to stimulatetheoretical and practical research in the field of parallel image processingand analysis. The volume contains three invited papers, a summary of a tutorial lecture, and twenty selected and refereed communications.

The research area of graph grammars and graph transformations dates back only two decades. But already methods and results from the area of graph transformation have been applied in many fields of computer science, such as formal language theory, pattern recognition and generation, compiler construction, software engineering, concurrent and distributed systems modelling, and database design and theory. This volume contains 24 selected and revised papers from an international seminar held in Dagstuhl, Germany, in 1993. The papers cover topics in the following areas: foundations of graph grammars and transformations; and applications of graph transformations to concurrent computing, specification and programming, and pattern generation and recognition.

This book contains the 61 papers that were accepted for presenta tion at the 1992 British Machine Vision Conference. Together they provide a snapshot of current machine vision research throughout the UK in 24 different institutions. There are also several papers from vision groups in the rest of Europe, North America and Australia. At the start of the book is an invited paper from the first keynote speaker, Robert Haralick. The quality of papers submitted to the conference was very high and the programme committee had a hard task selecting around half for presentation at the meeting and inclusion in these proceedings. It is a positive feature of the annual BMV A conference that the entire process from the submission deadline through to the conference itself and publication of the proceedings is completed in under 5 months. My thanks to members of the programme committee for their essential contribution to the success of the conference and to Roger Boyle, Charlie Brown, Nick Efford and Sue Nemes for their excellent local organisation and administration of the conference at the University of Leeds."

Sacred Geometry exists all around us in the natural world, from the unfurling of a rose bud to the pattern of a tortoise shell, the sub-atomic to the galactic. A pure expression of number and form, it is the language of creation and navigates the unseen dimensions beyond our three-dimensional reality. Since its discovery, humans have found many ways - stone circles, mandalas, labyrinths, temples- to call upon this universal law as a way of raising consciousness and communicating with a divine source. By becoming aware of the dots and lines that build the world around you, Sacred Geometry will teach you how to bring this mystical knowledge into your daily practice.

This monograph presents the author's studies in music recognition aimed at developing a computer system for automatic notation of performed music. The performance of such a system is supposed to be similar to that of speech recognition systems: acoustical data at the input and music scoreprinting at the output. The approach to pattern recognition employed is thatof artificial perception, based on self-organizing input data in order to segregate patterns before their identification by artificial intelligencemethods. The special merit of the approach is that it finds optimal representations of data instead of directly recognizing patterns.

This monograph is an outgrowth of the authors' recent research on the de velopment of algorithms for several low-level vision problems using artificial neural networks. Specific problems considered are static and motion stereo, computation of optical flow, and deblurring an image. From a mathematical point of view, these inverse problems are ill-posed according to Hadamard. Researchers in computer vision have taken the "regularization" approach to these problems, where one comes up with an appropriate energy or cost function and finds a minimum. Additional constraints such as smoothness, integrability of surfaces, and preservation of discontinuities are added to the cost function explicitly or implicitly. Depending on the nature of the inver sion to be performed and the constraints, the cost function could exhibit several minima. Optimization of such nonconvex functions can be quite involved. Although progress has been made in making techniques such as simulated annealing computationally more reasonable, it is our view that one can often find satisfactory solutions using deterministic optimization algorithms."

In this book a global shape model is developed and applied to the analysis of real pictures acquired with a visible light camera under varying conditions of optical degradation. Computational feasibility of the algorithms derived from this model is achieved by analytical means. The aim is to develop methods for image understanding based on structured restoration, for example automatic detection of abnormalities. We also want to find the limits of applicability of the algorithms. This is done by making the optical degradations more and more severe until the algorithms no longer succeed in their task. This computer experiment in pattern theory is one of several. The others, LEAVES, X-RAYS, and RANGE are described elsewhere. This book is suitable for an advanced undergraduate or graduate seminar in pattern theory, or as an accompanying book for applied probability, computer vision, or pattern recognition.

This volume contains the papers from the first British Neural Network Society meeting held at Queen Elizabeth Hall, King's College, London on 18--20 April 1990. The meeting was sponsored by the London Mathemati cal Society. The papers include introductory tutorial lectures, invited, and contributed papers. The invited contributions were given by experts from the United States, Finland, Denmark, Germany and the United Kingdom. The majority of the contributed papers came from workers in the United Kingdom. The first day was devoted to tutorials. Professor Stephen Grossberg was a guest speaker on the first day giving a thorough introduction to his Adaptive Resonance Theory of neural networks. Subsequent tutorials on the first day covered dynamical systems and neural networks, realistic neural modelling, pattern recognition using neural networks, and a review of hardware for neural network simulations. The contributed papers, given on the second day, demonstrated the breadth of interests of workers in the field. They covered topics in pattern recognition, multi-layer feedforward neural networks, network dynamics, memory and learning. The ordering of the papers in this volume is as they were given at the meeting. On the final day talks were given by Professor Kohonen (on self organising maps), Professor Kurten (on the dynamics of random and structured nets) and Professor Cotterill (on modelling the visual cortex). Dr A. Mayes presented a paper on various models for amnesia. The editors have taken the opportunity to include a paper of their own which was not presented at the meeting."

Mustererkennung heisst, ahnlich wie bei Sinneswahrnehmungen mit Sensoren Signale aus der technischen Umwelt zu empfangen und mit Hilfe zuvor gelernter Situationen momentane Messungen zu interpretieren und dabei im Hinblick auf neue Eindrucke lernfahig zu sein. Anlasslich des 11. DAGM-Symposiums wurden zu diesem Themenkomplex nahezu 100 Arbeiten eingereicht, von denen 42 Vortrage und 38 Plakatprasentationen zur Tagung und fur dieses Buch ausgewahlt wurden. Der Band enthalt Aufgabenstellungen, Denkweisen und neuere Forschungsergebnisse aus den Gebieten Mustererkennung, Bildverstehen, Bildfolgen, Wissensverarbeitung und Spracherkennung.

This book offers readers a broad view of research in some Western and Eastern European countries on pattern and signal analysis, and on coding, handling and measurement of images. It is a selection of refereed papers from two sources: first, a satellite conference within the biannual International Conference on Pattern Recognition held in Rome, November 14-17, 1988, and second, work done at the International Basic Laboratory on Image Processing and Computer Graphics, Berlin, GDR. The papers are grouped into three sections. The first section contains new proposals for the specific computation of particular features of digital images and the second section is devoted to the introduction and testing of general approaches to the solution of problems met in digital geometry, image coding, feature extraction and object classification. The third section illustrates some recent practical results obtained on real images specifically in character and speech recognition as well as in biomedicine. All the techniques illustrated in this book will find direct application in the near future. This book should interest and stimulate the reader, provoke new thoughts and encourage further research in this widely appealing field.

In this volume the author gives an introduction to the theory of group representations and their applications in image science. The main feature of the presentation is a systematic treatment of the invariance principle in image processing and pattern recognition with the help of group theoretical methods. The invariance properties of a problem often largely define the solution to the problem. Invariance principles are well known in theoretical physics but their use in image processing is only a few years old. The reader will find that group theory provides a unifying framework for many problems in image science. The volume is based on graduate-level lectures given by the author, and the book is intended for students and researchers interested in theoretical aspects of computer vision.

In this second edition every chapter of the first edition of Pattern Analysis has been updated and expanded. The general view of a system for pattern analysis and understanding has remained unchanged, but many details have been revised. A short account of light and sound has been added to the introduction, some normalization techniques and a basic introduction to morphological operations have been added to the second chapter. Chapter 3 has been expanded significantly by topics like motion, depth, and shape from shading; additional material has also been added to the already existing sections of this chapter. The old sections of Chap. 4 have been reorganized, a general view of the classification problem has been added and material provided to incorporate techniques of word and object recognition and to give a short account of some types of neural nets. Almost no changes have been made in Chap. 5. The part on representation of control structures in Chap. 6 has been shortened, a section on the judgement of results has been added. Chapter 7 has been rewritten almost completely; the section on formal grammars has been reduced, the sections on production systems, semantic networks, and knowledge acquisition have been expanded, and sections on logic and explanation added. The old Chaps. 8 and 9 have been omitted. In summary, the new edition is a thorough revision and extensive update of the first one taking into account the progress in the field during recent years.

Die 5. Osterreichische Artificial-Intelligence-Tagung setzt sich zusammen aus wissenschaftlichem Programm, Workshops und Tutorials. Der wissenschaftlich orientierte Teil des Tagungsprogramms umfasst sowohl eingeladene als auch begutachtete Vortrage zu den Themen Qualitatives Schliessen, Methodik Wissensbasierter Systeme und deren Anwendung, Logik/Deduktion, Naturlichsprachliche Systeme, Lernen und Kognition. Zum Informationsaustausch waren zusatzlich Workshops zur Weiterbildung vorgesehen. Besonders das Thema "Philosophie und KI" demonstrierte das allgemeine Interesse. Dies soll mit Beitragen dokumentiert werden, die einen Uberblick uber Beruhrungspunkte der KI mit philosophischen Stromungen bieten und auch den Einfluss der KI als Teil der Informatik auf das philosophische Weltbild verdeutlichen. Ebenfalls reprasentative Beitrage wurden zu den Workshops "Konnektionismus", "Qualitatives Schliessen" und "Begriffsbildung/-modellierung" ausgewahlt.

Pattern recognition is traditionally considered to cover all aspects of sensory data perception ranging from data acquisition, through preprocessing and low level analysis, to high level interpretation. Owing to its breadth and important application potential, the field of pattern recognition has been attracting considerable attention of researchers in academia and industry and consequently it has been witnessing a rapid growth and perpetual development. The need for dissemination of the latest results is being served by a host of international conferences on pattern recognition. One such series of meetings is regularly held in the United Kingdom under the auspices of the British Pattern Recognition Association. This volume contains papers presented at the BPRA 4th International Conference on Pattern Recognition held in Cambridge, March 28-30, 1988. Alongside the conventional topics of statistical and syntactic pattern recognition, contributions address issues in the hot subject areas of adaptive learning networks, computer vision, knowledge base methods and architectures for pattern processing, and among others, report progress in the application domains of document processing, speech and text recognition and shape analysis for industrial robotics. It is believed that the collection is not merely a report on current activities but that it will also be an important source of inspiration for future developments in the field of pattern recognition.

Software design patterns are known to play a vital role in enhancing the quality of software systems while reducing development time and cost. However, the use of these design patterns has also been known to introduce problems that can significantly reduce the stability, robustness, and reusability of software. This book introduces a new process for creating software design patterns that leads to highly stable, reusable, and cost-effective software. The basis of this new process is a topology of software patterns called knowledge maps. This book provides readers with a detailed view of the art and practice of creating meaningful knowledge maps. It demonstrates how to classify software patterns within knowledge maps according to their application rationale and nature. It provides readers with a clear methodology in the form of step-by-step guidelines, heuristics, and quality factors that simplify the process of creating knowledge maps. This book is designed to allow readers to master the basics of knowledge maps from their theoretical aspects to practical application. It begins with an overview of knowledge map concepts and moves on to knowledge map goals, capabilities, stable design patterns, development scenarios, and case studies. Each chapter of the book concludes with an open research issue, review questions, exercises, and a series of projects.

Das Buch fuhrt auf einfache und verstandliche Weise in die Bayes-Statistik ein. Ausgehend vom Bayes-Theorem werden die Schatzung unbekannter Parameter, die Festlegung von Konfidenzregionen fur die unbekannten Parameter und die Prufung von Hypothesen fur die Parameter abgeleitet. Angewendet werden die Verfahren fur die Parameterschatzung im linearen Modell, fur die Parameterschatzung, die sich robust gegenuber Ausreissern in den Beobachtungen verhalt, fur die Pradiktion und Filterung, die Varianz- und Kovarianzkomponentenschatzung und die Mustererkennung. Fur Entscheidungen in Systemen mit Unsicherheiten dienen Bayes-Netze. Lassen sich notwendige Integrale analytisch nicht losen, werden numerische Verfahren mit Hilfe von Zufallswerten eingesetzt."

Der vorliegende Tagungsband enthalt die meisten Vortrage der gemeinsamen Tagung der Deutschen Gesellschaft fUr Angewandte Optik (DGaO) und der Deutschen Arbeits gemeinschaft fUr Musterkennung (DAr, M). Die Beziehungen der beiden Organisationen sind folgendermaBen: Wahrend die DGaO eine traditionsreiche wissenschaftliche Gesellschaft ist, stellt die DAGM einen problemorientierten Dachverband folgender wissenschaftlicher Gesellschaften dar: - Deutsche Gesellschaft fUr angewandte Optik (DGaO) - Deutsche Gesellschaft fUr Nuklearmedizin (DGNM) - Deutsche Gesellschaft fUr Ortung und Navigation (DGON) - Deutsche Gesellschaft fUr Medizinische Dokumentation, Informatik und Statistik (GMDS) - Deutsche Gesellschaft fUr Angewandte Datenverarbeitung und Automation in der Medizin (GADAM) - Gesellschaft fUr Informatik (GI) - Nachrichtentechnische Gesellschaft (NTG). Die DAGM fordert den Erfahrungsaustausch auf dem Gebiet der Mustererkennung und ist Mitglied der International Association for Pattern Recognition (IAPR). Die verantwortungsvolle Arbeit des Programmausschusses wurde unter Vorsitz von Prof. Hertel, Berlin, von folgenden Herren geleistet: Dr. Dreher, Bonn; Prof. Fercher, Essen; Dr. Kossel, Wetzlar; Prof. Lanzl, Oberpfaffenhofen; Dipl.lng. Platzer, MUnchen; Dr.-Ing. Poppl, MUnchen; Dr. Pretschner, Hannover; Prof. Schmidt, Oberkochen; Dr. Triendl, Oberpfaffenhofen; Dr. Walter, MUnchen; Prof. Winkler, Berlin. Unser ganz besonderer Dank gilt dem SchriftfUhrer der Deutschen Gesellschaft fUr Angewandte Optik, Herrn PreuB, Wetzlar, insbesondere fUr seine Mitarbeit im ProgrammausschuB und fUr die Drucklegung des gemeinsamen Tagungsprogramms. Wir wUnschen allen Teilnehmern, daB diese gemeinsame Tagung fUr sie ein interessanter und fruchtbarer Erfahrungsaustausch wird."

Das Rahmenthema der 25. Jahrestagung der Deutschen Gesellschaft fur Medizinische DOkumentation, Informatik und Statistik e. V. - Nachsorge und Krankheitsverlaufsanalyse - hat einen engen Bezug zu aktuellen Problemen des Gesundheitswesens. Insbesondere in Kliniken, in denen schwere Erkrankungen mit modernen erfolgversprechenden Massnahmen be- handelt werden, wird die Notwendigkeit einer systematischen weiteren Uberwachung dieser Patienten immer dringlicher erachtet. Die Beob- achtung des weiteren Schicksals, die arztliche Betreuung und die Bewertung der zeitlichen Verlaufe ergeben Ansatzpunkte fur die weitere Verbesserung der Therapie. Diese Aufgabe lasst sich nur be- waltigen, wenn die dabei auftretenden Probleme der planvollen Doku- mentation, der Informationsubermittlung, der Datenspeicherung und der statistischen Auswertung von den Vertretern unseres Faches aktiv in Angriff genommen werden. Nach 25 Jahren einer sturmischen techno- logischen und Methodenentwicklung ist unsere junge medizinische Dis- ziplin in der Lage - wenn die erforderliche apparative und perso- nelle Ausstattung zur Verfugung steht -, die Probleme der Nachsorge und Krankheitsverlaufsanalyse in Zusammenarbeit mit Klinikern und Allgemeinmedizinern wirksam zu bearbeiten und Ergebnisse zu zeitigen, die fur den Arzt relevant sind. Die Tagung soll dazu Anregungen ver- mitteln und Losungswege aufzeigen. Funf Workshops und apparative, organisatorische und methodische Pro- bleme runden den Bezug unserer Arbeit auf die Probleme der !1edizin von heute ab. Erlangen hat eine traditionelle Verbundenheit mit der technologischen Entwicklung in der Medizin. Auch heute besteht, sowohl in der Medi- zinischen wie in der Technischen Fakultat, auf deren Campus wir tragen, ein waches Interesse fur den humanen Einsatz technologischer Neuerungen, insbesondere auch der elektronischen Datenverarbeitung. L.

The fundamental mathematical tools needed to understand machine learning include linear algebra, analytic geometry, matrix decompositions, vector calculus, optimization, probability and statistics. These topics are traditionally taught in disparate courses, making it hard for data science or computer science students, or professionals, to efficiently learn the mathematics. This self-contained textbook bridges the gap between mathematical and machine learning texts, introducing the mathematical concepts with a minimum of prerequisites. It uses these concepts to derive four central machine learning methods: linear regression, principal component analysis, Gaussian mixture models and support vector machines. For students and others with a mathematical background, these derivations provide a starting point to machine learning texts. For those learning the mathematics for the first time, the methods help build intuition and practical experience with applying mathematical concepts. Every chapter includes worked examples and exercises to test understanding. Programming tutorials are offered on the book's web site.