Extracting Silhouettes from 3D Models
三维模型轮廓线抽取算法

Silhouettes play a very important role in graphics communication. Because silhouettes are view dependent, it is expensive to extract silhouettes for previous techniques when the object is moving in space. This paper presents two algorithms of extracting silhouettes from 3D models. By exploiting the property of local maximum value and connectivity of silhouettes, the methods first identify part of silhouettes, then extract external silhouettes of 3D models by simple comparisons. The first method divides the model into some zones evenly in X and Y direction. Then it calculates the maximum edges of the zones in X and Y direction. These edges are silhouette edges. By searching the neighbor maximum edges of the silhouette edges, it finds the silhouette of the model. The second approach begins from the top edge in Y direction. By comparing the angles between the silhouette edge and its neighbor edges, it obtains the silhouette easily. Experimental results illustrate the efficiency of the methods. The algorithms and previous methods are compared and analyzed at the end of the paper.