等位函數法在影像切割之研究

Abstract

近年來，隨著電腦科技的發展，數位化影像的處理與應用漸趨重要。 在許多應用之前，影像切割是一個重要的基本處理。因此影像切割成為影像處理上的一個重要課題。本研究的主要目的是研究使用等位函數法(Level Set Methods)的動態曲線(Active Contour)影像切割法。等位函數法是一個新穎的數值方法，可以很容易的表示出曲線或曲面。同時也具有容易處理偏微分運算的特性。我們整理和實做使用等位函數法的一些基本且重要的影像切割法，並討論他們在實際運作上會遇到的問題及在影像切割上的優缺點。

In recent years, with the development of computer technology, digital image processing and its applications are more and more important. Image segmentation is an important and basic processing before other higher level applications. Image Segmentation has become one of the major subjects in image processing. The purpose of this thesis is to explore the image segmentation using Level Set Methods. Level Set Methods are novel numerical methods, they express curves or surfaces easily and have the Convenient feature of dealing with PDE calculations. We implement several basic and important methods of image segmentation using level set methods. Then discuss the main problems encountered in practice and their pros. and cons. on image segmentation.