摘要
为了有效地简化稠密采样点模型,提出了一种基于粒子群优化聚类算法的点模型简化方法.引入了具有强大全局寻优能力的粒子群优化算法,对传统的k-均值聚类算法进行改进,基于改进的聚类算法对点模型进行简化,选取具有最优个体适应度函数的粒子作为原始采样点集的最终简化模型.算法聚类依据采样点的空间位置、法向和曲率的邻近性,实现了点模型表面区域几何特征保持的简化.同时在聚类区域的划分中考虑了曲率阈值和区域半径,使得算法在有效地保持特征边界和曲面细节的同时,能够生成高质量的简化曲面.实验结果表明,粒子群优化的k-均值聚类算法克服了传统聚类算法容易陷入局部极小的缺点,具有更好的全局收敛性和较快的收敛速度.该简化方法在有效简化点模型的同时,很好地保持了原始模型的几何形状,且在相同简化效率下能够生成更高质量的简化曲面.
To efficiently simplify the densely sampled point model,a point sample data reduction method based on Particle Swarm Optimization algorithm（KPSO） was proposed,which was introduced to improve the traditional k-means clustering,and the improved clustering algorithm was used to simplify point model and select the best fitness value substitutes for original data points to simplify the model.The original point model was clustered into clusters according to the position,normal and curvature neighborhood.The clustering region partition according to the curvature threshold and regional radius not only can effectively preserve the feature edges and the surface details,but also can achieve the simplified point set generating high quality surface approximations to the origin point set surfaces.
出处
《海南大学学报（自然科学版）》
CAS
2010年第3期241-247,共7页
Natural Science Journal of Hainan University
基金
2010年广东省自然科学基金项目(10451064101005826)
广州市科技攻关计划项目(B07B2070870)
2007年广州市科技攻关计划项目(B07B2070870)
科技部科技型中小型企业技术创新基金无偿资助项目(02C26214400224)
广东省科技计划资助项目(2002A1020104)
关键词
点模型简化
粒子群算法
特征保持
K-均值聚类算法
point sampled surfaces simplification
Particle Swarm Optimization algorithm
preserve feature
k-means algorithm
