研究了用圆弧曲率估算截面数据离散曲率的方法和特征点提取方法，提出了基于曲线拟合的截面特征识别方法。
Other preprocessing methods such as discrete curvature estimation and feature point extraction are also studied, and a section-feature recognition algorithm, based on curve fitting, is proposed.
应用实例表明，算法的适用于任何测量点云，提取的边界数据精度高，可直接用于曲线拟合。
Empirical results show that the proposed algorithm is suitable to any measured data and the extracting boundary data is precise which can be used to fit curve directly.
应用推荐