3-D Block-Rooting Scheme with Application to Medical Image Enhancement

Abstract

The goal of image enhancement is to improve certain features and details on an image. It is a critical procedure in image processing (image segmentation, feature extraction), medical imaging (such as X-ray, CT, and MRI) and computer vision (target, object, and text detection; segmentation, registration, and recognition). X-ray, CT, and MRI images are often affected by blurriness and lack of contrast, but the clarity of these images is very important for the accuracy of medical diagnosis and treatment. This article offers (1) a 3-D transform block-rooting scheme-based image enhancement method and (2) a non-reference transform-domain quality measure to choose the presented algorithm parameters optimally. Experimental results from NYU, fastMRI, and ChestX-ray data sets show that the proposed technique performs well and can reduce noise during the sharpening of the image details. At the same time, it can undoubtedly be used in different medical image processing systems.