"Shape Representation and Similarity for Image Databases",
 
 Call for Papers:
 Special Issue of Pattern Recognition on
 
 "Shape Representation and Similarity for Image Databases"
 
 The recent increased use of image and multimedia databases has motivated
 research in developing and applying shape similarity measures.  In a
 sense this continues the long history of work in both shape
 representation and similarity.  However, since in image databases the
 object classes are generally unknown a priori, a global, non-parametric
 shape representation is often necessary.  By this we mean approaches in
 which the entire object shape is described uniformly in accord with our
 visual perception.  Such a representation should permit the
 differentiation of perceptually similar objects that are not
 mathematically identical.  This involves the use of shape representation
 based on visual parts, multiscale contour approximations, Hausdorff
 distances, skeleton based representations, curvature scale spaces,
 reaction-diffusion spaces, and deformable shapes. 
 
 Methods based on vectors of shape parameters like area, perimeter,
 elongation (major axis / minor axis), etc. are excluded.  Vectors of
 shape parameters may be very useful for shape classification, but not as
 a basis for shape similarity measures, since in order to recognize
 common shapes, hundreds of parameters must be represented explicitly,
 and most of these parameters are probably unknown.  Such a
 representation also excludes parametric methods that describe certain
 object classes, e.g., B-spline surface patches of car prototypes, since
 they require explicit assumptions about the category of objects to be
 represented. 
 
 We seek papers on global, non-parametric shape representations and shape
 similarity measures that are useful tools for retrieval of similar
 objects in image data bases and for other tasks related to image
 indexing, e.g., finding repeated patterns and symmetry groups.  Papers
 on comparison and analysis of different approaches with respect to shape
 related tasks in image databases are also welcome.  For example, what
 representations are most functional for a system that needs to reason
 about shape, find similar shapes fast under a given set of allowable
 transformations, or determine whether a given shape satisfies a set of
 user-related queries. 
 
 Please submit five copies of a manuscript to one of the guest editors.
 The deadline for submission is June 30, 1999.
 
 Guest Editors:
    Ari Gross, Department of Computer Science
       Queens College, Flushing, NY 11367, USA
       Email: ari@vision.cs.qc.edu
    Longin Jan Lateckim, Department of Applied Mathematics
       University of Hamburg, Bundesstr. 55, 20146 Hamburg, Germany
       Email: latecki@math.uni-hamburg.de
    Robert Melter, Department of Mathematics
       Long Island University, Southampton, NY 11968, USA
       Email: RMelter@aol.com