一种保细节特征的点云多尺度三角网格重建算法

A multi-scale triangular mesh reconstruction algorithm for detailed features preserving

点云三角网格重建是逆向工程领域重要的研究内容。网格模型的细节特征是评判三角网格重建算法的重要指标,如何通过点云数据重建完整的网格模型并保证模型表面的细节特征不丢失是点云重建的难点。为了解决这一问题,引入多尺度概念,在点云数据的低频尺度空间中进行网格重建,保证点云数据的重建率,从而保证有高质量的重建结果。在进行网格重建过程中,采用自适应半径搜索策略进行邻域点搜索,减少了网格模型的孔洞。实验结果表明,所提算法三角网格重建质量较高,具有较高的点云数据重建率,有效保证了重建结果细节特征的完整性。

Triangular mesh reconstruction of point cloud is an important research in reverse-engineering field.The preservation of detail features is an index to evaluate the triangular mesh reconstruction algorithm.How to reconstruct a complete mesh model from a point cloud and ensure that the detailed features of the model are not lost is difficult.To solve the problem,this paper introduces the concept of multi-scale to transfer point cloud from high frequency scale space to low frequency scale space.The multi-scale operation can guarantee the rate of reconstruction,which means good detailed features preserving.At the same time,we adopt the search strategy of adaptive radius to search neighborhood points,which can reduce holes of the result.The experiments show that the results of our method are of high quality,high reconstruction rate,and of good detailed features preserving.