Image Texture Classification Based on Immune Particle Swarm Optimization

LI Lin-yi,LI De-ren

DOI: https://doi.org/10.3321/j.issn:1001-1595.2008.02.010

2008-01-01

Abstract:Particle swarm optimization is a stochastic global optimization method based on swarm intelligence.It was developed through simulation of a simplified social system such as bird flocking.In order to improve the convergence performance of standard particle swarm optimization,immune particle swarm optimization is proposed by means of introducing memory ability and diversity of biologic immune system into standard particle swarm optimization in this paper.The particles in the swarm are constructed and swarm search strategies of immune particle swarm optimization are proposed in terms of the needs of the classification application after many texture features such as Laws texture energy template features and wavelet features are extracted from texture samples.Then the image texture classification algorithm based on immune particle swarm optimization is implemented to solve image texture classification problems.Compared with standard particle swarm optimization in the image texture classification experiment,immune particle swarm optimization has better convergence performance in training the image texture samples to get the centers of classification and the classification method based on it has higher classification accuracy.