Image enhancement using hybrid GSA — Particle swarm optimization

In this paper, a new approach for image enhancement of gray-level images using gravitational search algorithm and particle swarm optimization (PSO) is represented. The equation for the updates of velocity and position of particles in PSO is modified. For that, the values of two variables, iteration IT and α are evaluated through a newly developed Fuzzy Inference System. In PSO, each iteration updated the velocity of the particle, the velocity is dependent on the acceleration of the particle, which in turn is dependent on force applied, this force is optimized using Newton's law of gravity and motion. This makes the convergence of the PSO to yield better result as compared to the classical PSO, when applied for the enhancement of the images. The intensity component is enhanced using the unsharp masking technique. A new contrast gain is defined for amplification of the unsharp mask to produce the enhanced image. The saturation component is enhanced using the power-law transformation. A new objective function comprising entropy, image exposure, histogram flatness and histogram spread is introduced and optimized using PSO to learn the parameters used for the enhancement of a given image. The proposed approach is evaluated using different test images. Different performance measures are used for the quantitative analysis of the proposed approach.