| Popis: |
We consider graph theoretical approach to image processing, especially to segmentation of the biological images. At first, we provide a brief overview of methods from graph theory used for to image segmentation. Next, we focus on graph cuts method and its application in image processing. We show how we transform an image to a graph and corresponding segmentation network. Then, we show how we deal with segmentation of the real biological data. We define a completely new method for pre-processing data to get realistic results. We bring new pre-processing algorithm for preparing data for segmentation and also postprocessing counting of results and verifying the hypotheses. The chapter fullfils expectations of medical and biological researchers to get mathematical methods and tools for analysing the results and verifying the hypotheses. It refers to user-defined requirements of data processing, numerical evaluation, counting of cells or minerals, percentual performing, counting of minerals in the tissue. Our own contribution is the “software”, is created on special properties, requirements and requests on biological and medical researchers, because free and open software commonly available on Internet does not give them satisfying results. We want to show that graphs and its applications bring enough good tools for processing iron in medical and biological data. We used biological data in the connection with Alzheimer disease to achieve our goal. |
