Optimized Haar wavelet-based blood cell image denoising with improved multiverse optimization algorithm

Abstract

ABSTRACT Blood cells are complicated in the field of biomedicine. Several facilities now employ microscopic images to detect cells or parasites by technicians. But there is no medical device which can acquire the medical images without adding the noise. Hence, denoising the image is a vital process before predicting the illnesses for getting accurate results. In recent years, denoising the medical images by wavelet transform (WT) plays a major part in image processing. The major problem that occurs in denoising the medical images is to uphold the image quality of the denoised image. For that, a novel Optimised Haar WT integrated with Improved Multiverse Optimisation Algorithm based on Adaptive Thresholding function (IMVO-AT) is introduced to denoise the medical blood cell images. The proposed denoising method utilizes Haar WT (HWT), Improved Multiverse Optimisation Algorithm based on Adaptive Thresholding function (IMVO-AT), trilateral filter and an Inverse HWT (IHWT) to obtain the denoised blood cell image. Finally, the performance is evaluated in terms of PSNR, MSE, SMAPE, NC, SSIM, MAE, PC and RMSE for the proposed denoising approachand other denoising approaches. The experimental result depicts that the proposed denoising approach overcome all the other methods in terms of outcome and performance.