【作者】 黄锦红；

【导师】 邵东向；

【作者基本信息】 哈尔滨工业大学 ， 机械制造及其自动化， 2009， 硕士

【摘要】 图像的三维重建在很多领域具有重要的意义,本文研究基于CCD图像的三维重建的目的,就是利用从CCD传感器上获得的序列微观形体的图像切片,经过图像处理后提取出微观形体的轮廓线,利用合适的算法重建出形体的三维模型,这可以使微观形体能够方便的在空间坐标内进行定性或定量的分析。具体来说,本文的三维重建系统分四个步骤来完成,一是图像的预处理及轮廓线的提取。采用基于边缘检测的方法对图像进行分割,凸显出要重建物体的边缘轮廓,然后用轮廓线跟踪法提取轮廓线的坐标点集。二是轮廓线间的插值。为了减小轮廓线层间距离与层内像素间距离的差距,使其尽量能构造出均匀的三维数据场,就必须在每层轮廓线之间插入新的轮廓线层。按等角度对轮廓线进行取点,然后用线性插值的方法进行插值,这种方法简单易行。三是轮廓线间的三角划分,即通过一定的算法规则,将相邻轮廓线上的点上下连接,形成互不相交的三角形,从而构成物体的三维表面。本文用最短对角线法进行三角划分。最后是三维显示,即在二维屏幕上显示出重建后的三维体。在VC平台下结合OpenGL编写程序进行三维显示,并通过鼠标来控制图形的平移、缩放和旋转等操作。更多还原

【Abstract】 The 3-D reconstruction of image is of great significance in various domains. We discuss the technology of the 3-D reconstruction based on the CCD image in this paper. According to the characteristics of CCD image , First of all we pretreat the images collected: smoothing of the images in order to weaken the impact of noise; detecting of the image edge and extracting the contour in order to help the 3-D reconstruction next. Then use the shortest diagonal method to triangulate the sectional image.Specifically,The system of 3-D reconstruction is composed of four steps. The first step is getting the edge contours of object that will be reconstructed from the 2-D images. Several arithmetic operator is used to segment the images. After that the contours of prostate are very clear. Then the contours are outlines and the coordinate datasets are obtained by tracking contour method.The second step is interpolating some new contours between two contours witch is next to each other. Because the distance of contours at z-axis is bigger than the distance of contours at x-axis and y-axis, it is needed to create serial new contours between original contours to form the symmetrical 3-D data field. It is the method that gets the contour’s points by an equal angle and then interpolates with linearity interpolation algorithm.The third step is plotting triangles, which is the algorithm of reconstructing 3-Dobject’s surface by linking the points of two contours to become a serial of triangles without intersecting each other. Method of the shortest diagonal algorithm is used.The last step is 3-D display in computer. We can show the 3-D object with the combination of VC platform and OpenGL. The 3-D object can be translated and rotated by mouse conveniently.更多还原