从特征提取和特征匹配两方面考虑，提出了一种鲁棒的形状匹配方法。首先，基于求和不变量，设计了基于面积的形状参数化和归一化方法，提出了参数化求和不变量，该不变量基于形状局部描述且采用积分算子计算，具有较好的鲁棒性和仿射不变性。然后，为进一步提高形状匹配的鲁棒性，在特征匹配上，分析了参数化求和不变量的先验信息，设计了基于特征重整的匹配距离函数，并通过动态规划进行实现。仿真实验表明了所提方法的有效性。

A robust shape matching method is presented in this paper. An area based parameterization technique is designed for shape resampling. Then the parameterized summation invariant (PSI) is generated based on the resampled shape and summation invariant (SI). PSI which uses local parts of a shape to characterize the shape and can be calculated via integral operation is affine invariant and robust against noise. In order to improve the robustness of shape matching, prior information of PSI is analyzed to define a notion of distance between shapes based on feature regulating. The proposed distance measure can be implemented by dynamic programming. Experimental results show that the proposed method is valid.