scientific article; zbMATH DE number 194137
Thirty two papers on pattern recognition and artificial intelligence are published in the three parts of the book: I. Image Processing (Section I: Image Processing Systems, Section II: Image Processing Algorithms). II. Pattern Recognition. III. Artificial Intelligence and Pattern Recognition. In the first section of part I, including three papers, image analysis is distinguished from image processing. The Acuity software package for quantitative analysis of complex images is proposed by I. Young, R. Roos. A hierarchically structured software package for image processing is presented by P. Dewaele et al. The last paper by E. Person deals with the basic requirements for real-time image analysis with respect to industrial applications. In Section II, Part I, ten papers are concerned with image processing algorithms. Finding lines, edges, shape and symmetry features are subject to the first group of papers. P. E. Danielsson and Q. Z. Ye describe rotation- invariant operators for detecting the magnitude and orientation of lines, with application to fingerprint recognition. An edge detection scheme is discussed by L. J. van Vliet and coauthors: the non-linear Laplace operator and the Marr-Hildreth model of edge detection are used for detection of one pixel thick edges. The symmetry is considered by J. Bigün as a pair of conjugate harmonic functions; modeling symmetries of the neighbourhoods in grey value images is based on the form of the iso- grey value curves. The paper by A. Smeulders and M. Worring is concerned with a method for computing accurate, anisotropic local shape measures from digitized patterns. A survey of some algorithms for computing visibility relations between edges of a simple polygon is presented by G. Toussaint. Applications of image processing in robotics are subject to the next two papers. A closed form solution for locating objects in 3D space is proposed by G. Nagy and A. Falsafi. K. Rijnierse and F. Groen give a technique for robot stereovision system. The final three papers discuss applications of image processing in medicine. A 3D reconstruction of the myocardial perfusion from biplane angiograms is proposed by A. Dumay et al. P. J. H. van Guyck and coauthors develope a technique for automated detection of the centerline in a coronary angiogram. H. Stark and H. Peng consider the estimation of the location, size and shape of opacities in computer tomograms. Seven papers are presented in Part II. They are concerned with classification, methodology and applications for pattern recognition. The design and performance of the Parzen window classifier is investigated by A. Jain and M. Ramaswami. The bootstrap criterion function is proposed and compared with other estimation techniques. Discriminant analysis based on a learning set of objects is described by R. Duin and E. Backer in terms of the possibilities. Fuzzy labels are introduced, parametric and nonparametric classification schemes are given. The basic methods of classification using incomplete data sets are presented and analyzed by C. Queiros and E. Gelsema. Estimation of the missing values in continuous and discrete feature spaces are proposed, some simulating experiments are discussed. S. Morgera and A. Soleymani present a model for the pattern recognition theory, based on a generalization of rate-distortion theory. Some mapping methods for multidimensional pattern analysis are described by W. Siedlecki et al. The least square mapping is used in experiments in cluster analysis and the design of piecewise linear classifiers. The last two papers are concerned with methodology. Classification of speciments in automated cytometry is studied by T. Timmers and E. Gelsema. The model allows both between- and within-specimen variability of the features and is applicable to a cell-class approach as well as to a cell-feature approach. Astronomical object classification is described by M. Kurtz. The last eleven papers, organized as Part III, are about artificial intelligence and pattern recognition. The paper by M. Musen and J. van der Lei shows the challenges in creation of expert and pattern recognition systems. It discusses the ways for both artificial intelligence and pattern recognition communities in developing computational models for classification tasks. R. Bhatnagar and L. Kanal describe a graph based formalism that represents a link between observable variables and variables of interest in a given problem situation characterized by incomplete and/or uncertain knowledge. Five well-known techniques dealing with uncertainty in the reasoning, the ascerts and the role of uncertainty, are analyzed by A. M. van Ginneken and A. Smeulders. Combining independent classifiers based on the Dempster-Shafer theory is proposed by E. Mandler and J. Schürmann. A knowledge base system for approximate reasoning in exploratory data analysis is presented by E. Backer. A decision method to detect pattern evolution is described by C. Cocca and B. Dubuisson. The next four papers are devoted to pattern recognition and artificial intelligence technique in bio-medical applications: cerebral blood vessels, molecular biology, chromosome aberration scoring, analysis of biological systems. In the last paper, by B. Dawant and B. Jansen, a coupled expert system for signal understanding is described. Discussions at the conference time are included.
This page was built for publication:
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4040929)
