中文摘要：

针对标准的最小二乘拟合方法计算球面调和展开时延展性差且计算大尺度线性方程组时精度低的问题，文中给出一个迭代残量拟合算法，用来求解含有实值球面调和函数的大尺度线性方程组，能够精确地生成大尺度3D曲面的球面调和模型。算法易于实现，对硬件要求低。

英文摘要：

The standard least square fitting （LSF） method for spherical harmonics expansion is not scalable and cannot accurately solve the large-scale linear systems .An iterative residual fitting （IRF ） algorithm is proposed for the large-scale linear systems containing spherical harmonics ,with its effectiveness and scalability demonstrated in creating accurate SH models for large 3D surfaces .IRF is very easy to implement and requires few machine resources .