根据已有实物的测量数据进行模型重建 ,在机械产品逆向建模、计算机视觉、基于二维轮廓数据的生物外形重建等领域中具有重要应用价值。随着坐标测量设备的发展 ,获取包含被测物体更多细节的海量数据已非常方便 ,但大量的测量点却给模型重建带来了困难。本文首先提出了精度可控的海量数据自动简化算法。为了提高算法的效率 ,文中提出了一个数据集空间划分策略。根据简化后的数据集 ,应用步进立方体方法重建模型的三角网格曲面表示。由于种种原因 ,重建的三角网格模型常常含有不希望有的孔洞。为此 ,本文给出了一个算法产生形状优化的三角片以修补网格模型中的孔洞。经过孔洞修补 ,完全封闭的三角网格模型可以直接输出为快速原型制造中广泛应用的 STL文件。应用实例说明了本文的方法的可行性 Reconstruction of the model based on existing physical measurement data has important application value in the fields of reverse modeling of mechanical products, computer vision, reconstruction of biological appearance based on two-dimensional contour data. With the development of coordinate measuring equipment, it has been very convenient to obtain a large amount of data that contains more details of the measured object. However, a large number of measurement points have caused difficulties in model reconstruction. This paper first proposed an automatic control algorithm for massive data with controllable precision. In order to improve the efficiency of the algorithm, a data set space partitioning strategy is proposed. According to the simplified data set, a triangular mesh surface representation of the model is reconstructed using the step cube method. For various reasons, the reconstructed triangular mesh model often contains unwanted holes. To this end, this paper presents an algorithm to generate shape-optimized triangular patches to repair holes in the mesh model. After hole repair, a fully enclosed triangular mesh model can be exported directly to STL files widely used in rapid prototyping. The application example shows the feasibility of this method