Theoretical Research, Based On The Complexity Of The Visual Surface Reconstruction

Complex surface reconstruction based on vision has received much attention in the Computer Vision area in recent years. In this field, a number of different approaches to surface height reconstruction from a single image (shape from shading) are analyzed and discussed in this paper. Based on these approaches, with the theory of minimization of functional, an iteration method of surface orientation and surface height reconstruction from a single image is concluded and with Gauss-Newton iteration, nonlinear term is converted into a linear term. The concrete steps solving all above are shown in the paperThe algorithm program is developed with Visual C++ Programming Language on the platform of Windows95 operation system After that, data are transported to CAXA ME2000 system, CAD/CAM software. Finally, by machining a model on CNC system, the algorithm is verified and the result is satisfying...