International Journal of Computer Applications |
Foundation of Computer Science (FCS), NY, USA |
Volume 70 - Number 25 |
Year of Publication: 2013 |
Authors: Nalin Goel, K. K. Shukla |
10.5120/12222-8371 |
Nalin Goel, K. K. Shukla . Modified Evolutionary Autonomous Agents Approach to Image Feature Extraction. International Journal of Computer Applications. 70, 25 ( May 2013), 13-18. DOI=10.5120/12222-8371
In this paper, we present a modified evolutionary autonomous agents method for extracting features from a grey scale image. Features can be extracted on the basis of differences in intensity values of the target pixel with the neighboring pixels. We use agents to identify such change in the intensity values with the neighborhood. Agents are single unit entity with the power of computation and behaviors. The agent possess behaviors such as Self-reproduction, Diffusion and Death. Self-reproduction occurs when the parent agent detects the feature lying beneath it. Self-reproduction results in generation of other agents. Diffusion to another pixel occurs in case the former condition is not met until the agent attains certain age and it finally Dies. Self-reproduction and diffusion occurs in the direction governed by the previous state of the agent. Agents are independent entities hence the power of parallelizability can be really advantageous.