Enhanced Medical Images Through Multi-Scale Mathematical Morphology By Reconstruction

Medical images are indispensable tools for several medical tasks, by allowing a much more accurate diagnosis to help in decision making. For this reason it is essential to have good quality medical images. However, sometimes they show degradations such as poor contrast or imperfections in details. In this article we present an algorithm that improves the details of medical images, preserves the average brightness, preserves the structural similarity and corrects the problem of poor contrast. This algorithm improves medical images using the top-hat transform by reconstruction, which extracts brightness and darkness features on multiple scales. These features are used to enhance the medical image. The algorithm was tested with medical images from two public databases. Experimental results show that the proposed algorithm improves contrast, introduces little noise, preserves natural brightness, detail and similarity to the original medical image.