中文摘要：

为使匹配结果收敛于全局最优解，提出基于统计特征的点云模型匹配技术。通过调整自由模型与固定模型的一个或多个对应统计特征重合或一致来实现模型的匹配。基于统计特征的模型匹配分为完全匹配和部分匹配两种情况。完全匹配能够一次匹配成功；部分匹配需要进一步处理才能完成模型匹配：交互调整自由模型中未被约束的自由度，使得自由模型与固定模型达到视觉上的匹配，然后采用迭代最近点算法精确匹配。根据统计特征进行模型匹配，使匹配方案稳定、可靠，保证匹配结果收敛于全局最优解。

英文摘要：

To make matching result converge to global optimal solution, a point cloud model matching technology based on statistical features was proposed. The model matching was realized through keeping corresponding statistical features of free model and fixed model consistent. The statistical- features-based matching was classified as： complete matching and segment matching. Complete matching could be matched without the second try, but the segment matching needed follow-up handling to finish model matching： interactively adjust unconstrained degrees of freedom to acquire the visually matching of free model and fixed model, and then accurately match them by using Iterative Closest Point （ICP） algorithm. The matching scheme which used statistical features was stable and reliable, and it also guaranteed that the matching result was converged to global optimal solution.