Use of aggregation pheromone density for image segmentation
Pattern Recognition Letters
| Hi-index | 0.01 |
Social insects like ants, bees deposit pheromone (a type ofchemical) in order to communicate between the members of theircommunity. Pheromone, that causes clumping behavior in a species andbrings individuals into a closer proximity, is called aggregationpheromone. This article presents a new algorithm (called, APC) forpattern classification based on the property of aggregationpheromone found in natural behavior of real ants. Here each datapattern is considered as an ant, and the training patterns (ants)form several groups or colonies depending on the number of classespresent in the data set. A new (test pattern) ant will move alongthe direction where average aggregation pheromone density (at thelocation of the new ant) formed due to each colony of ants is higherand hence eventually it will join that colony. Thus each individualtest ant will finally join a particular colony. The proposedalgorithm is evaluated with a number of benchmark data sets in termsof classification accuracy. Results are compared with other state ofthe art techniques. Experimental results show the potentiality ofthe proposed algorithm.