基于表面的角膜三维重建及其在临床眼科医学中的应用

【作者】 杨朋翠；

【导师】 初学导；

【作者基本信息】 曲阜师范大学 ， 控制理论及控制工程， 2004， 硕士

【摘要】 二维图像的三维重建是医学影像学的研究热点之一。作为一种新兴的技术，三维重建利用二维图像的有用信息，立体显示医学图像，提高了图像的直观性，更有利于医学和科技工作者对所需要的信息进行分析和显示。 论文首先设计三维重建所遵循的一般步骤，在各步骤中，介绍实现各具体功能的常用方法，比较各种方法的优缺点，然后集中实现角膜切片的三维重建。论文以计算机为处理工具，借助Visual C++编程环境，运用数字图像处理、计算机图形学、眼科学等理论知识，借鉴其他相关论文的优点、特点和方法，设计简洁易懂、可读性和可操作性强的程序算法，进行切片重组。 文章针对角膜切片成像的特殊性及特点，根据切片在三维空间的位置，用立体几何知识，得出二维切片边缘像素的三维位置坐标。采用B-rep造型，将角膜切片的表面划分成若干小区域，每个小区域用四边形小面来逼近，使用小面来构造三维模型。对于一系列的三维像素点数据，可认为其为四边形面元的顶点，按一定的规律对顶点进行定义，再用同样的方法对面元进行定义，从而建立起角膜切片的三维模型。利用这种思想，可以对角膜进行建模，这是本论文着重论述的部分。 在三维重建的过程中，实现了一组图片的自动打开功能，结合阈值方法和多值图像数学形态学方法快速准确地提取切片边缘的像素信息，从而进行基于表面的三维重建。在程序设计中，考虑了显示的直观性和灵活性，兼顾了美观性，通过友好的人机界面，实现了角膜图像的多角度观察。更多还原

【Abstract】 3D Reconstruction of 2D image is one of hotspots of medical image research. As a new technology, 3D Reconstruction utilizes useful messages of 2D images to show medical image tridimensionally. It improves the visibility of the image, which enables medical and scientific people to analyze or display what they need.At first, the article frames the steps that 3D Reconstruction follows. In the process, we introduce common methods which are used to realize various functions and compare the characters of them. Then we focus our effort to accomplish the 3D Reconstruction of cornea slices. The article utilizes Visual C++, digital image management, computer graphics and ophthalmological theoretical knowledges, uses the merits, characters and methods of other correlative articles for reference. It projects compact, readable and operable procedure steps to accomplish the reconstruction of the slices.The article aims at the particularities and characters of cornea slices. It bases on the positions of three dimension and utilizes the knowledges of solid geometry to get the 3D positional coordinates of 2D slices’ marginal elements. The procedure adopts B-rep sculpt to divide the surface of cornea to many small parts. Each small part is simulated by small rectangle. So integrate the small rectangles to form 3D model. With a series of 3D element data, we consider them as the vertexes of small rectangles and define them according certain rule. In the same way, we define the facets. At last, we can build up the 3D Model of cornea slices. This is the main part the article emphasizes on.In the process of 3D Reconstruction, we achieve the automatic opening function of a group of images. With threshold value according with multiple-value mathematical morphological method, we pick up the marginal elements of slices accurately . The procedure accomplishes the 3D reconstructionof a group of cornea sheets successfully based on surface. In the course of the design of the procedure, we accomplish the cornea image from various points of view, considering the visibility, promptness and beauty of the display.更多还原