Optimised DWT using cooperative particle swarm optimiser for hybrid domain based medical and natural image denoising

The quest for productive image denoising systems still is a valid challenge, at the intersection of practical investigation and measurements. In spite of the sophistication of the recently proposed systems, most calculations have not yet achieved an attractive level of applicability. In this research, an optimal wavelet filter coefficient design-based methodology is proposed for image denoising. The method utilises new wavelet filter whose coefficients are derived by discrete wavelet (Haar) transform using CPSO optimisation and bilateral filter. The optimal wavelet coefficient based denoising methods minimise the noise, while bilateral filter further decreases the noise and increases the PSNR without any loss of relevant image information. Overall, the proposed approach consists of two stages namely, (i) design of optimal wavelet filter, (ii) image denoising using a bilateral filter. At first, wavelet optimal coefficients are selected using cooperative particle swarm optimiser (CPSO). After that, the hybrid domain based algorithm (wavelet with bilateral filter) is applied to the noisy image which is helpful to obtain the denoised image. A comparative study of the performance of different existing approaches and the proposed denoised approach is made in terms of PSNR, SDME, SSIM and GP. When compared, the proposed algorithm gives better PSNR compared to the existing methods.