A novel Approach of Retinal Image Enhancement using PSO System and Measure of Fuzziness

Abstract

Retinal image analysis has become a promising research field in the diagnosis of ophthalmological diseases since optic disc, fovea, and blood vessels are essential features based on which the analysis is performed. This paper introduces a novel retinal image enhancement scheme based on particle swarm optimization (PSO) under fuzzy framework as it overcomes the problem of upgrading the imprecise nature of retinal image in traditional PSO system in the field of image enhancement. Firstly, two sub-images such as lower and upper in fuzzy regions of the image are determined by type-2 fuzzy system. Then,an S-shape membership function is utilized on lower and upper sub-images. An objective function such as index of fuzziness is optimized in terms of particle swarm optimization (PSO) system and it also defines the adaptive parameters which are incorporated with the proposed enhancement technique. The suggested model applies low contrast color retinal images to treat different diseases of eye such as diabetic retinopathy, glaucoma, optic disc etc. The proposed algorithm demonstrates excellent result in terms of both subjective and objective evaluation than traditional PSO systems and other baseline methods.