Local search-based dynamically adapted bat algorithm in image enhancement domain

Abstract

Bat algorithm (BA) is a new metaheuristic optimisation algorithm, which has already proved its supreme performance on many optimisation fields. However, it is possible to increase its efficiency when solving complex optimisation problems. This study concentrates on improving the efficiency of BA by incorporating different types of local search strategies and novel self-adaption strategies of parameters such as loudness, pulse rate and frequency. Comparative analysis of three different proposed local search strategies has been performed to find the best one. The proposed modified BAs with local search strategies are employed to solve five popular image enhancement models. Experimental results prove that self-adaption of parameters enhances the capability of standard BA. But the addition of efficient local search technique with self-adaption increases the effectiveness of the standard BA to a great extent.