An Adaptive Bacterial Foraging Algorithm for color image enhancement

Abstract

A new approach in spatial domain is presented for the enhancement of color images using an Adaptive Bacterial Foraging Algorithm (ABFA). The image may be classified as under-exposed or over-exposed depending upon the value of exposure. A new objective function is formulated which makes use of the fuzzy entropy. This objective function is optimized using ABFA which allows the step-size of the bacterial colony to vary dynamically over the generations. The lifetime of the bacterial colony is described by generations which are split into two phases; exploration and exploitation based on the value of step-size. Smaller step-sizes correspond to exploitation phases in which the entire bacterial colony is trying to exploit a region of interest while larger step-sizes correspond to exploration phases in which the entire bacterial colony is trying to explore the search space to find regions of interest which can then be exploited by reducing the step-size. This method is applicable on both over-exposed and under-exposed images. The proposed algorithm is found to be better and much more simplified than the existing Bacterial Foraging Algorithm (BFA) in fuzzy domain for color image enhancement upon comparison.