摘要
使用成长型神经网络(GCS)三维建模,需要不断分裂网格中计数器值最大的节点;本文根据新增节点的Voronoi面积大小分配其计数器值。实验结果表明:该算法对曲面重建具有较好的效果。
Using Growing Cell Structures to realize surface reconstruction, the node whose signal counter is the largest need be splitted continually. The paper distributes a signal counter to a node by calculating the area of new Voronoi cells Some experiment results are given which show that the method presented is effective in reconstruction.
出处
《安徽建筑工业学院学报(自然科学版)》
2007年第5期86-88,共3页
Journal of Anhui Institute of Architecture(Natural Science)
基金
安徽省教育厅自然科学项目(KJ2007B311ZC)
安徽建筑工业学院硕士启动基金(2006070124)