The Study Of 3D Reconstruction Method Based On Image Sequences

The geometric modeling is very important in science research and engineering applications. The modeling based on intensity images is a new method in modeling, it is simple in modeling and convenient in operating in contrast to traditional methods, so it will be used in wide areas. A whole procedure of reconstructing geometric shape and modeling from image sequences is presented. The procedure is made up of three parts: the first part is camera calibration, the second is acquiring discrete points ,the last is presenting the surface of object. First , the camera calibration theory , one method and its example are simply presented reconstruction object from image sequences which is the main part of the article is presented in succession. This part is made up of two parts, one is acquiring discrete points using voxel coloring~method , the method uses the object抯 color sufficiently , it can reconstruct many kinds of objects including occluding object, the method avoids image correspondence problems by working in discretized scene space whose voxels are traversed in a fixed visibility ordering , so it has fast speed. The method using NURBS curves fitting discrete points and then using NURBS surface skinning NURBS curves are presented to represent the surface of object. The algorithm of Delaunay triangulation for a set of 3D points is presented.