作者: V.B. Nitya , N. Sridevi , Arun K. Pujari
DOI: 10.1016/0167-8655(92)90128-M
关键词:
摘要: Abstract Recovery of 3D shape from the silhouettes object has gained interest among researchers in computer vision. Volume intersection is underlying paradigm wherein reconstructed by intersecting volumes generated sweeping along respective viewing directions. Several attempts have been made to design efficient algorithms for recovery using this paradigm. These differ one another representation schemes, perspective versus orthographic projection and An algorithm proposed earlier generate a linear octree its three mutually orthogonal views. Generally, there some degree any real life situation are unrealistic. In paper we propose an projections. The makes use novel technique segment tree structure achieve efficiency. Experimental results also reported.