IASC-CI: Improved Ant Based Swarm Computing for Classifying Imagery


The social insect metaphor for working out arrays of predicaments has become promising vicinity in latest years focusing on indirect or direct interactions among various agents. Swarm Computing has a multidisciplinary character as its study provides insights that can help humans manage complex systems that offer an alternative way of designing intelligent systems which emphasises on the emergent collective intelligence of groups of simple agents. Classification is the computational procedure [3] [5] that sort the images into groups according to their similarities. Numerous methods for classification have been developed and exploring new methods to increase classification accuracy has been a key topic. Ant Colony Optimization (ACO) [2] [6] is an algorithm inspired by the foraging behaviour of ants wherein ants leaves the volatile substance called pheromone on the soil surface for the purpose of foraging and collective interaction via indirect communications. This paper focuses on improved Methodology of Swarm Computing for classifying imagery exploring various techniques such as IASC (Improved Ant based Swarm Computing), DWT [15] (Discrete Wavelet Transform) and SVM [1] (Support Vectors Machines) for edge detection [14][15][19], feature extraction[10], feature selection [12], and finally image classification [5][6]. KeywordsClassification, Imagery, Feature Extraction, Feature Selection, Pheromone, Swarm Computing, Discrete Wavelet Transform, State Vector Machine, Improved Ant based Swarm Computing, Ant Colony Optimization, Edge Detection, Image Classification.

6 Figures and Tables

Cite this paper

@inproceedings{Rai2014IASCCIIA, title={IASC-CI: Improved Ant Based Swarm Computing for Classifying Imagery}, author={Rebika Rai and Ratika Pradhan}, year={2014} }