An Efficient Algorithm for Computing Multi-scale Connectivity Measures

Authors:
Georgios K. Ouzounis
Affiliations:
Second Department of Surgery School of Medicine, Democritus University of Thrace, University General Hospital of Alexandroupoli, Alexandroupoli, Greece 68100
Venue:
ISMM '09 Proceedings of the 9th International Symposium on Mathematical Morphology and Its Application to Signal and Image Processing
Year:
2009

Citing 9
Cited 0

Set-Theoretical Algebraic Approaches to Connectivityin Continuous or Digital Spaces

Journal of Mathematical Imaging and Vision
Connectivity on Complete Lattices

Journal of Mathematical Imaging and Vision
Connected morphological operators for binary images

Computer Vision and Image Understanding
Shape Connectivity: Multiscale Analysis and Application to Generalized Granulometries

Journal of Mathematical Imaging and Vision
Connectivity on complete lattices: new results

Computer Vision and Image Understanding
Connected Shape-Size Pattern Spectra for Rotation and Scale-Invariant Classification of Gray-Scale Images

IEEE Transactions on Pattern Analysis and Machine Intelligence
Mask-Based Second-Generation Connectivity and Attribute Filters

IEEE Transactions on Pattern Analysis and Machine Intelligence
Antiextensive connected operators for image and sequence processing

IEEE Transactions on Image Processing
Flat zones filtering, connected operators, and filters by reconstruction

IEEE Transactions on Image Processing

Quantified Score

Hi-index	0.00

Visualization

Abstract

Multi-scale connectivity measures have been introduced in the context of shape analysis and image segmentation. They are computed by progressive shape decomposition of binary images. This paper presents an efficient method to compute them based on the dual-input Max-Tree algorithm. Instead of handling a stack of binary images, one for each scale, the new method reads a single gray-level image, with each level associated to a unique scale. This reduces the component labeling iterations from a total number equal to the number of scales to just a single pass of the image. Moreover, it prevents the repetitive decomposition of each component under study, for the remaining scale range, since these information are already mapped from the input image to the tree hierarchy. Synthetic and real image examples are given and performance issues are discussed.