Shape Classification And Analysis ─ Theory And Practice

Because the properties of objects are largely determined by their geometric features, shape analysis and classification are essential to almost every applied scientific and technological area. A detailed understanding of the geometrical features of real-world entities (e.g., molecules, organs, materials and components) can provide important clues about their origin and function. When properly and carefully applied, shape analysis offers an exceedingly rich potential to yield useful applications in diverse areas ranging from material sciences to biology and neuroscience.

Get Access to the Authors’ Own Cutting-Edge Open-Source Software Projects—and Then Actually Contribute to Them Yourself!

The authors of Shape Analysis and Classification: Theory and Practice, Second Edition have improved the bestselling first edition by updating the tremendous progress in the field. This exceptionally accessible book presents the most advanced imaging techniques used for analyzing general biological shapes, such as those of cells, tissues, organs, and organisms. It implements numerous corrections and improvements—many of which were suggested by readers of the first edition—to optimize understanding and create what can truly be called an interactive learning experience.

New Material in This Second Edition Addresses



Graph and complex networks


Dimensionality reduction


Structural pattern recognition


Shape representation using graphs


Graphically reformulated, this edition updates equations, figures, and references, as well as slides that will be useful in related courses and general discussion. Like the popular first edition, this text is applicable to many fields and certain to become a favored addition to any library.

Visit http://www.vision.ime.usp.br/~cesar/shape/ for Useful Software, Databases, and Videos

Luciano da Fontoura Costa is a full professor at the Instituto de Física de São Carlos, University of Sao Paulo, Brazil, and head of the Cybernetic Vision Research Group and the Multidisciplinary Computing Group. He holds BScs in Electronic Engineering and Computer Science, as well as a MSc in Applied Physics and a PhD in Electronic Engineering. Roberto M. Cesar Jr. holds a BSc in Computer Science, a M.Sc. in Electrical Engineering, and a Ph.D in Computational Physics. He is currently a lecturer at the University of Sao Paulo.

INTRODUCTION
INTRODUCTION TO SHAPE ANALYSIS
CASE STUDIES
COMPUTATIONAL SHAPE ANALYSIS
ADDITIONAL MATERIAL
ORGANIZATION OF THE BOOK


BASIC MATHEMATICAL CONCEPTS
BASIC CONCEPTS
LINEAR ALGEBRA
DIFFERENTIAL GEOMETRY
MULTIVARIATE CALCULUS
CONVOLUTION AND CORRELATION
PROBABILITY AND STATISTICS
FOURIER ANALYSIS
GRAPHS AND COMPLEX NETWORKS


SHAPE ACQUISITION AND PROCESSING
IMAGE REPRESENTATION
IMAGE PROCESSING AND FILTERING
IMAGE SEGMENTATION: EDGE DETECTION
IMAGE SEGMENTATION: ADDITIONAL ALGORITHMS
BINARY MATHEMATICAL MORPHOLOGY
FURTHER IMAGE PROCESSING
REFERENCES


SHAPE CONCEPTS
INTRODUCTION TO TWO-DIMENSIONAL SHAPES
CONTINUOUS TWO-DIMENSIONAL SHAPES
PLANAR SHAPE TRANSFORMATIONS
CHARACTERIZING 2D SHAPES IN TERMS OF FEATURES
CLASSIFYING 2D SHAPES
REPRESENTING 2D SHAPES
SHAPE OPERATIONS
SHAPE METRICS
MORPHIC TRANSFORMATIONS


SHAPE REPRESENTATION
INTRODUCTION
PARAMETRIC CONTOURS
SETS OF CONTOUR POINTS
CURVE APPROXIMATIONS
DIGITAL STRAIGHT LINES
HOUGH TRANSFORMS
EXACT DILATIONS
DISTANCE TRANSFORMS
EXACT DISTANCE TRANSFORM THROUGH EXACT DILATIONS
VORONOI TESSELLATIONS
SCALE-SPACE SKELETONIZATION
BOUNDING REGIONS


SHAPE CHARACTERIZATION
STATISTICS FOR SHAPE DESCRIPTORS
SOME GENERAL DESCRIPTORS
FRACTAL GEOMETRY AND COMPLEXITY DESCRIPTORS
CURVATURE
FOURIER DESCRIPTORS


MULTISCALE SHAPE CHARACTERIZATION
MULTISCALE TRANSFORMS
FOURIER-BASED MULTISCALE CURVATURE
WAVELET-BASED MULTISCALE CONTOUR ANALYSIS
MULTISCALE ENERGIES


SHAPE RECOGNITION
INTRODUCTION TO SHAPE CLASSIFICATION
SUPERVISED PATTERN CLASSIFICATION
UNSUPERVISED CLASSIFICATION AND CLUSTERING
A CASE STUDY: LEAVES CLASSIFICATION
EVALUATING CLASSIFICATION METHODS


STRUCTURAL SHAPE RECOGNITION
INTRODUCTION
SYNTACTIC PATTERN RECOGNITION
REGION DECOMPOSITION
GRAPH MODELS
SPATIAL RELATIONS
GRAPH MATCHING
CASE STUDY: INTERACTIVE IMAGE SEGMENTATION
COMPLEX NETWORKS FOR IMAGE AND SHAPE ANALYSIS


EPILOGUE
FUTURE TRENDS IN SHAPE ANALYSIS AND CLASSIFICATION