针对基于最小二乘法的ICP 曲面匹配算法难以处理待比较曲面的局部大变形问题, 提出一种改进算

法。即采用遗传算法确定曲面初始相对位置以保证匹配优化结果为全局最优值, 利用ICP 算法匹配结果构造

偏差阈值, 以此阈值过滤点群后再以最小二乘法进行匹配处理, 消除局部大变形影响, 获得合理的变换矩阵。以此变换矩阵变换初始点群再进行误差计算, 从而获得理想的匹配结果-Least square method based on the ICP surface matching algorithm to be relatively difficult to deal with the local surface deformation problem, an improved algorithm. Genetic algorithm that is used to determine the relative position of the initial surface in order to ensure optimization of the results of matching the global optimum value, the use of ICP algorithm to match the results of structural deviation threshold, this threshold point group and then filtered to match the least squares processing, the elimination of local large deformation effects of the transformation matrix to obtain reasonable. Transformation matrix to transform the initial point of the calculation error of another group in order to achieve an ideal match