位置:成果数据库 > 期刊 > 期刊详情页
基于区域生长的网格模型分割技术
  • 期刊名称:计算机辅助设计与图形学学报,2006,18(7):1011-1015
  • 时间:0
  • 分类:TP391[自动化与计算机技术—计算机应用技术;自动化与计算机技术—计算机科学与技术]
  • 作者机构:[1]哈尔滨工业大学计算机学院,哈尔滨150001, [2]东北农业大学工程学院,哈尔滨150036
  • 相关基金:国防科技预研项目;国家自然科学基金(60075010,60475011)
  • 相关项目:最优小波自适应图像压缩的关键技术研究
中文摘要:

针对三角形网格模型简化中保持细节特征的要求,提出了以顶点和三角形作为种子的区域生长策略,并采用可变阈值的生长方法.首先将与顶点相关的三角形面积的均值作为顶点的属性值,在区域生长过程中重复选择具有极大属性值的图元作为种子;在预分割后,区域要进行合并和平滑处理.实验表明,该方法快速有效,能够满足网格模型简化中保持局部区域特征的需要.

英文摘要:

Focusing on detail preserving in triangle mesh model simplification, matching of the local average size of meshes surrounding a vertex or a triangle is taken as the criterion of region growth, and the matching threshold is lowering down in proportion to the increment of mesh density in a region. At first, the area of each mesh and their average value surrounding each vertex are computed and the process of region growth is started repetitively from a seed of the largest mesh. After completion of preliminary segmentation, merging of compatible separate regions and smoothing of their boundaries are undertaken. Experiment shows that this approach is fast and effective, meeting the requirement of keeping local features in mesh simplification.

同期刊论文项目
同项目期刊论文