An Efficient Cascaded Filter Design using Dragonfly Algorithm for Image Noise Reduction

Abstract

This paper describes an optimal image denoising method with an efficient cascaded filter structure designed using dragonfly algorithm. Though there are several conventional filtering methods used for image denoising, the proposed method shows much improved result in terms of PSNR, IQI and SSIM values keeping the entire image attributes intact. This proposed image denoising technique exhibits its effectiveness not only in the matter of both quantitative and visual aspects of image but also the performance shows accuracy in presence of various types of noise like Gaussian, Salt and Pepper, and Speckle with different variance values. Furthermore, the experimental results with different real images establish the fact that this approach achieves better optimal solution than existing denoising techniques.