Pattern Recognition Theory and Applications: Proceedings of the NATO Advanced Study Institute Held at St. Anne's College, Oxford, March 29-April 10, 1

ISBN-13: 9789027713797 / Angielski / Twarda / 1982 / 576 str.

J. Kittler; K. S. Fu; L. F. Pau

Opis

This book is the outcome of the successful NATO Advanced Study Institute on Pattern Recognition Theory and Applications, held at St. Anne's College, Oxford, in April 1981., The aim of the meeting was to review the recent advances in the theory of pattern recognition and to assess its current and future practical potential. The theme of the Institute - the decision making aspects of pattern recognition with the emphasis on the novel hybrid approaches - and its scope - a high level tutorial coverage of pattern recognition methodologies counterpointed with contrib- uted papers on advanced theoretical topics and applications - are faithfully reflected by the volume. The material is divided into five sections: 1. Methodology 2. Image Understanding and Interpretation 3. Medical Applications 4. Speech Processing and Other Applications 5. Panel Discussions. The first section covers a broad spectrum of pattern recognition methodologies, including geometric, statistical, fuzzy set, syntactic, graph-theoretic and hybrid approaches. Its cove, r- age of hybrid methods places the volume in a unique position among existing books on pattern recognition. The second section provides an extensive treatment of the topical problem of image understanding from both the artificial intelligence and pattern recognition points of view. The two application sections demonstrate the usefulness of the novel methodologies in traditional pattern 'recognition application areas. They address the problems of hardware/software implementation and of algorithm robustness, flexibility and general reliability. The final section reports on a panel discussion held during the Institute.

Spis treści

Section 1 Pattern Recognition Methodology.- Advances in Nonparametric Techniques of Statistical Pattern Classification.- Statistical Pattern Recognition of Binary Variables.- On the Haar Condition in Algorithms for the Optimal Solution of Linear Inequalities.- Pattern Class Representation Related to the Karhunen-Loeve Expansion.- Sample Size Sensitive Entropy.- An Optimal Clustering Method Based on the Rasson’s Criterion and Resulting from a New Approach.- Computational Geometric Problems in Pattern Recognition.- A Coding Approach to Pattern Recognition.- Fuzzy Set Theoretic Approach: A Tool for Speech and Image Recognition.- Syntactic Methods in Pattern Recognition.- Hybrid Approaches to Pattern Recognition.- Pattern Recognition of Relational Structures.- Hybrid Pattern Recognition: A Synthesis of the Numerical and Qualitative Approaches.- Fusion of Multisensor Data in Pattern Recognition.- Section 2 Image Understanding and Interpretation.- Artificial Intelligence Approaches to Image Understanding.- Summary of the DARPA Image Understanding Research Program.- Three-Dimensional Object Representation.- Automatic Interpretation of Lines and Text in Circuit Diagrams.- A Modular System for Interpreting Binary Pixel Representations of Line-Structured Data.- Recognition of Time Varying Patterns in Image Sequences.- Section 3 Medical Applications.- Application of Statistical and Syntactical Methods of Analysis and Classification to Chromosome Data.- Attributes for Pattern Recognition Selected by Stepwise Data Compression Supervised by Visual Classification.- Application of Pattern Recognition to Knowledge System Design and Diagnostic Inference.- Applications of Interactive Pattern Recognition.- An Electrocardiographic Signal Understanding System.- Segmentation-Based Boundary Extraction for Endocardial Definition in Cross-Sectional Echocardiography.- Computerised Anatomy Atlas of the Human Brain.- Section 4 Speech Processing and Other Applications.- Decision Theoretic and Structural Pattern Recognition Methods in Speech Processing.- Pattern Recognition and Machine Intelligence Techniques for Electric Power System Security Assessment.- A Multi-Microcomputer Based Reading Machine for the Blind.- Decision-Set Concept with Application to the Recognition of Hand-Printed Numeric Characters.- On the Development of a Model for Determining the Frequency of Occurrence of English Language Words.- Section 5 Panel Discussion.- Summary of Panel Discussion on Decision Theoretic Methods.- Author Index.