中文摘要：

提出了一种以型值点集合描述自由曲面的数字化模型，据此设计了从密集数据点群重构曲面数字化模型的自组织拟合及重构算法．该算法利用局部规则对测量点群进行自组织拟合，形成型值点集合及局部规则的表达形式，每一个局部区域的型值点集合满足特定的拟合规则，造型和加工时根据局部规则进行插值和刀具轨迹计算．通过仿真对基于二次多项式局部规则的高次曲面的迭代拟合算法进行了分析，以一个实际的人脸模型数据为例，实现了从测量点群到自组织模型的重构和加工，并将其与SDSCAD1．0曲面重构系统进行对比，结果表明：计算简单且稳定性好，编程效率高．

英文摘要：

A digital model which represents free form curve with data points set is presented in this paper, self organization fitting and reconstructing algorithm which reconstructs digital model of curve from dense data points cluster was designed. Self-organization fitting based on local rules was applied to measurement points cluster in the algorithm, a expression method based on data points set and local rules is presented. Data points set within each local area meet some special fitting rules. Interpolation and tool-path calculation is according to local rules in model and machining. Iteration fitting algorithm of high-order curve based on local rules expressed by second order polynomial was analyzed by simulation. Actual model data of human-face was acted as an example, the process of reconstruction from measurement points to self-organization model and machining was achieved, it is compared with SDSCAD 1.0 curve reconstruction system, the result is better. The result shows the calculation is simple and stable, program efficiency is high.