Detekce objektů v medicínských snímcích

Abstract

The aim of this bachelor thesis is image segmentation. The theoretical analysis is consequently followed by its processing in software environment MATLAB (version 2011a). One of the main parts of image processing is the edge detection and therefore this work focuses especially on edge detection operators, namely those using approximation of the derivative of the first order, second order and operators approximating image function. The thesis deals with convolution which forms a necessary part image segmentation. It also contains the suggestion and realization of available algorithms implementation in MATLAB for the image edge detection and hereby helps the user to recognize objects of interest in individual medical images easily. The principal objective of this work is to implement own edge detection operators into a part of algorithms in order to find individual objects in images. The thesis also includes practical output of medical image data processing, specifically a method of detecting blood vessel and optic nerveusing by Kirsch operator.