无序多视角点云的自主配准方法

An Automatic Approach for Multi-View Registration of Unordered Range Scans

为了提高自主多视角点云配准方法的效率和精度,提出一种基于特征匹配的无序多视角点云全局配准方法,通过计算和匹配点云的特征描述子快速实现双视角点云配准;设计了有效的判定准则用于判别双视角配准的结果是否可靠;利用所提出的模型扩展方法对可靠的双视角配准结果进行点云模型的扩展。通过交替地执行双视角配准、配准结果判别和模型扩展,该方法可实现无序多视角点云的全局配准。在斯坦福图形学实验室公开数据集上的实验结果表明,与效果较优的同类方法相比,该方法可使得配准效率平均提高近5倍,且配准误差显著下降,同时可提高多视角点云配准的性能。

A feature match based approach is proposed registration~ This approach can achieve global registrat and matches feature descriptors to achieve pair wise reg judge the reliability of pair wise registration results is d method is proposed to use reliable results of pair wise to improve the performance ion of unordered range scans istrations. In addition, an ef esigned. Moreover, a model of multi view .It calculates fective rule to registration to augment the object mo Multi view registrations are accomplished by alternately executing the pair wise registrations judgments, and model augmentation.Experimental results on available public data sets del. and and comparisons with some state of the art approaches show that the proposed approach increases the registration efficiency by about 5 times on average and largely reduces registration errors.