An optimized histogram equalization with morphological feature enhancement and adaptive sized residual fusion network based segmentation in bio medical images

Abstract

To provide accurate and quick diagnoses nowadays, doctors or newlineradiologists need the finest medical image quality. The presence of noise newlinethroughout the capture, transmission, reception, storage, and retrieval phases newlineof medical images is often a concern. Medical images are distorted and of newlineworse quality when there is noise present. This degradation includes the newlinesuppression of edges, structural features, boundary-blurring, etc. Researchers newlineare still having difficulty removing noise from the original image. The newlinemajority of researchers created efficient noise-removal methods. The most newlinebasic, much researched, and mostly unresolved issue is the method of image newlinedenoising or restoration. This research may be split into two separate parts. newlineIn the first stage of the study at the moment, the fractional newlinedifferential method is the most successful method that may be used to deal newlinewith factual issues. For the purpose of denoising medical images, this newlineprojected approach makes use of the fractional derivatives definitions newline Riemann- Liouville (R-L), Grunwald-Letnikov (G-L), and the Caputo newlinetechnique . The suggested method is based on fractional derivative, which newlineenhances the overall quality of the image. The input image is then processed newlineusing an integer order approach, which may include pre-processing, image newlineconversion, and noisy image processing . The Riemann, Liouville, and newlineCaputo algorithms are used to carry out the process of applying the fractional newlinedifferential mask technique. After the noise in the medical image has been newlineremoved using an anisotropic diffusion method, the improved image is next newlineassessed, and eventually, a denoised and forecasted image is produced. newline