Binary Structuring Elements Decomposition Based on an Improved Recursive Dilation-Union Model and RSAPSO Method

Yudong Zhang,Yi Sun,Zhengchao Dong,Preetha Phillips,Shuihua Wang,Genlin Ji

doi:10.1155/2014/272496

Abstract

This paper proposed an improved approach to decompose structuring elements of an arbitrary shape. For the model of this method, we use an improved dilation-union model, adding a new termination criterion, as the sum of 3-by-3 matrix should be less than 5. Next for the algorithm of this method, we introduced in the restarted simulated annealing particle swarm optimization method. The experiments demonstrate that our method can find better results than Park's method, Anelli's method, Shih's SGA method, and Zhang's MFSGA method. Besides, our method gave the best decomposition tree of different SE shapes including “ship,” “car,” “heart,” “umbrella,” “vase,” “tree,” “cat,” “V,” “bomb,” and “cup.”

Highlights

Mathematical morphology (MM) is a theory and technique for the analysis and processing of geometrical structures based on set theory, lattice theory, topology, and random functions
We introduced the restarted simulated annealing (RSA) technique to improve the performance of Simulated annealing (SA)
The structuring element (SE) decomposition problem was first transformed into an optimization problem with virtue of the improved recursive dilation-union model that contains a new termination criterion

Summary

Introduction

Mathematical morphology (MM) is a theory and technique for the analysis and processing of geometrical structures based on set theory, lattice theory, topology, and random functions. MM is commonly applied to digital images, and can be employed to graphs, surface meshes, solids, and many other spatial structures [1]. In these applications, the inherent strategy in MM is to explore the characteristics of an object by probing its microstructure with various forms, known as structuring element (SE). Most image processing architectures adapted to morphological operations use SEs of a limited size. The techniques for decomposing a largesized SE into combined small-sized SEs are of importance [2]

Methods

Results

Conclusion