Variable Length PSO-Based Image Clustering for Image Denoising

View Sample PDF

Author(s): Somnath Mukhopadhyay (Calcutta Business School, India), J. K. Mandal (University of Kalyani, India)and Tandra Pal (National Institute of Technology Durgapur, India)
Copyright: 2016
Pages: 27
Source title: Handbook of Research on Natural Computing for Optimization Problems
Source Author(s)/Editor(s): Jyotsna Kumar Mandal (University of Kalyani, India), Somnath Mukhopadhyay (Calcutta Business School, India)and Tandra Pal (National Institute of Technology Durgapur, India)
DOI: 10.4018/978-1-5225-0058-2.ch013

Keywords: Artificial Intelligence / Biologically-Inspired Computing / Information Science Reference / Medicine & Healthcare

Purchase

View Variable Length PSO-Based Image Clustering for Image Denoising on the publisher's website for pricing and purchasing information.

Abstract

This chapter proposed a variable length Particle Swarm Optimization based image clustering technique for restoration of noises from digital images. Here in this two step noise restoration technique the noise free pixels are kept unchanged. The denoising technique uses 3 × 3 test window on the center pixel of the noisy image. Prior to detection and filtering, variable length PSO based image clustering has been done. The output of clustering determines the performance of the subsequent stages of the algorithm. For denoising weighted median filtering technique is proposed. Variable length particles are considered and randomly encoded for the initial population. The length of particles is changed by adding and/or deleting cluster centers present in the particles. Three evaluation criteria are used in the fitness function of the proposed algorithm. The performance of the proposed algorithm is compared with some similar algorithms existing in the literature on several standard digital images.

The IRMA Community

Research IRM

Variable Length PSO-Based Image Clustering for Image Denoising

Purchase

Abstract

Related Content

IRMA Sponsors