Segmentation is an essential method and crucial process to accurately analyze medical images.
Abstract
Segmentation is an essential method and crucial process to accurately analyze medical images.
For example, good separations between hard and soft tissues are necessary for analyzing medical imaging of the brain accurately.
We analyze a new segmentation method applied on medical images based on the definition of entropy. It helps to distinguish between tissues with different characteristics of medical imaging.
We examined the entropy formula to fit the segmentation of images: fit the formula to work with images and selecting the appropriate coefficients through experiments.
To simulate working with images from natural environments, we added noise to existing images, and examined the smoothness and clarity of the images after the entropy process.
First, we converted the images to gray scale mode, and examined noised images and images without noise. We applied the new method on 3 types of linear noise, and one non linear noise type.
Results
original image poisson noise
X-direction Entropy Y-direction Entropy
