Robust Denoising Technique for Ultrasound Images Based on Weighted Nuclear Norm Minimization

Abstract

Image denoising is an efficacious pre-processing requisite procedure of ultrasound image investigation. In this study two denoising techniques adopted and evaluated to compare their performance. The widespread use of ultrasound images facilitates the diagnosis of various diseases. They pose several challenges and hence efficient pre-processing pipelines are essential to extract useful diagnostic information from the images. Much light is thrown on the Common carotid artery (CCA) images in this study. Two approaches are endorsed for image denoising involving and converting to grayscale for effective diagnosis. Weighted nuclear norm minimization (WNNM) approach is found to be more impressive and better. This also bolstered the validation methods computed in the work. It pretends that the study is useful in extracting diagnostic information. The experimental results impart authenticity to the proposed technique in the adequate analysis of ultrasound images. The principle objective of this work is to aid and accentuate the succeeding processing stages such as segmentation and object recognition to facilitate accurate and exact diagnosis.