Graph Based Representations in Pattern Recognition

1. Front Matter

   Pages I-VIII

   PDF

2. Data Structures and Representation

   1. Construction of Combinatorial Pyramids

      • Luc Brun, Walter G. Kropatsch

      Pages 1-12

   2. On Graphs with Unique Node Labels

      • Peter J. Dickinson, Horst Bunke, Arek Dadej, Miro Kraetzl

      Pages 13-23

   3. Constructing Stochastic Pyramids by MIDES — Maximal Independent Directed Edge Set

      • Yll Haxhimusa, Roland Glantz, Walter G. Kropatsch

      Pages 24-34

3. Segmentation

   1. Functional Modeling of Structured Images

      • Jocelyn Marchadier, Walter G. Kropatsch

      Pages 35-46

   2. Building of Symbolic Hierarchical Graphs for Feature Extraction

      • Mickaël Melki, Jean-Michel Jolion

      Pages 47-58

   3. Comparison and Convergence of Two Topological Models for 3D Image Segmentation

      • Achille Braquelaire, Guillaume Damiand, Jean-Philippe Domenger, Frédéric Vidil

      Pages 59-70

4. Graph Edit Distance

   1. Tree Edit Distance from Information Theory

      • Andrea Torsello, Edwin R. Hancock

      Pages 71-82

   2. Self-Organizing Graph Edit Distance

      • Michel Neuhaus, Horst Bunke

      Pages 83-94

   3. Graph Edit Distance with Node Splitting and Merging, and Its Application to Diatom Identification

      • R. Ambauen, S. Fischer, Horst Bunke

      Pages 95-106

5. Graph Matching

   1. Orthonormal Kernel Kronecker Product Graph Mdatching

      • Barend Jacobus van Wyk, Michaël Antonie van Wyk

      Pages 107-117

   2. Theoretical Analysis and Experimental Comparison of Graph Matching Algorithms for Database Filtering

      • Christophe Irniger, Horst Bunke

      Pages 118-129

   3. A Comparison of Three Maximum Common Subgraph Algorithms on a Large Database of Labeled Graphs

      • D. Conte, C. Guidobaldi, C. Sansone

      Pages 130-141

   4. Swap Strategies for Graph Matching

      • Paolo Fosser, Roland Glantz, Marco Locatelli, Marcello Pelillo

      Pages 142-153

6. Matrix Methods

   1. Graph Matching Using Spectral Seriation and String Edit Distance

      • Antonio Robles-Kelly, Edwin R. Hancock

      Pages 154-165

   2. Graph Polynomials, Principal Pivoting, and Maximum Independent Sets

      • Roland Glantz, Marcello Pelillo

      Pages 166-177

   3. Graph Partition for Matching

      • Huaijun Qiu, Edwin R. Hancock

      Pages 178-189

7. Graph Clustering

   1. Spectral Clustering of Graphs

      • Bin Luo, Richard C. Wilson, Edwin R. Hancock

      Pages 190-201

   2. Comparison of Distance Measures for Graph-Based Clustering of Documents

      • Adam Schenker, Mark Last, Horst Bunke, Abraham Kandel

      Pages 202-213

   3. Some Experiments on Clustering a Set of Strings

      • Jean-Michel Jolion

      Pages 214-224

This volume contains the papers presented at the Fourth IAPR Workshop on Graph Based Representations in Pattern Recognition. The workshop was held at the King’s Manor in York, England between 30 June and 2nd July 2003. The previous workshops in the series were held in Lyon, France (1997), Haindorf, Austria (1999), and Ischia, Italy (2001). The city of York provided an interesting venue for the meeting. It has been said that the history of York is the history of England. There have been both Roman and Viking episodes. For instance, Constantine was proclaimed emperor in York. The city has also been a major seat of ecclesiastical power and was also involved in the development of the railways in the nineteenth century. Much of York’s history is evidenced by its buildings, and the King’s Manor is one of the most important and attractive of these. Originally part of the Abbey, after the dissolution of the monasteries by Henry VIII, the building became a center of government for the Tudors and the Stuarts (who stayed here regularly on their journeys between London and Edinburgh), serving as the headquarters of the Council of the North until it was disbanded in 1561. The building became part of the University of York at its foundation in 1963. The papers in the workshop span the topics of representation, segmentation, graph-matching, graph edit-distance, matrix and spectral methods, and gra- clustering.

• 3D
• algorithms
• calculus
• classification
• computational graph theory
• data structures
• database
• graph matching
• graph-based methods
• graph-theoretic methods
• information theory
• modeling
• object recognition
• pattern analysis
• pattern recognition

• Department of Computer Science, University of York, York, UK

  Edwin Hancock

• D.I.I.I.E., Università degli Studi di Salerno, Fisciano (SA), Italy

  Mario Vento

Policies and ethics