基于邻域漂移的点云法向估计算法研究

Point cloud normal estimation based on neighborhood drift

随着三维激光扫描仪的改进和普及,获取三维点云数据的方式越来越方便.法向量作为点云数据不可或缺的属性之一,在诸多算法中具有重要作用.由于受到噪声、离群点、非均匀采样等因素的影响,准确快速估计尖锐特征点的法向量仍然是具有挑战性的.提出基于邻域漂移的点云法向估计算法,实现准确快速地对尖锐特征点的法向进行估计.首先,对当前点的近邻点构造其邻域,所有近邻点所对应的邻域构成候选邻域集.利用协方差分析对候选邻域进行评价,并选取最优邻域用于最终的法向估计.实验结果表明本算法在法向估计的质量上与前沿算法持平,在运行速度上与传统PCA算法相近,可以最大程度兼顾法向质量与计算速度.

With the improvement and popularization of 3D laser scanner,it is more and more convenient to obtain 3D point cloud data.As one of the indispensable attributes of point cloud data,normal vector plays an important role in many algorithms.However,fast and reliable normal estimation is still an intriguing challenge since sharp features,noise,and sampling anisotropy are inevitably contained in acquired point clouds.In this paper,a point cloud normal direction estimation algorithm based on neighborhood drift is proposed,which can accurately and quickly estimate the normal direction of sharp feature points.Firstly,the neighborhood of the current point is constructed,and all the neighborhood corresponding to the current point constitutes the candidate neighborhood set.The candidated neighborhood is evaluated by covariance analysis and the optimal neighborhood is selected for the final normal estimation.The experimental results show that the proposed algorithm is equal to the cutting-edge algorithm in terms of the quality of the normal estimation,and similar to the traditional PCA algorithm in terms of the running speed,which can give consideration to both the normal quality and the computational speed to the greatest extent.