利用双边滤波法进行点云去噪的试验与分析

利用无人机影像生成点云数据时，会产生一些噪声点，这些噪声点对点云数据的分类提取等会带来不利影响。为了消除这些噪声点，利用栅格法去除浮在点云上方的稀疏点，并利用pcl库的StatisticalOutlierRemoval滤波器去除离群噪声点，以及利用双边滤波算法对小型的噪声点进行纠正。试验结果表明，本文方法既保留了目标地物的几何特征，又能够有效地去除影响目标地物识别的噪声。

Experiment and Analysis of Point Cloud Denoising Using Bilateral Filtering Method

In this paper, we extracted point cloud data from image of UAV with overlapping regions. There are some noise points while extracting the data. These noise points will have a great impact on the image classification and other operations. Therefore, we need to remove the noise points. A grid method is used to remove the sparse points above the point cloud. StatisticalOutlierRemoval filter of PCL Library is used to remove outliers noise. A bilateral filter is used to deal with small noise. Through experiments, we can not only retain the characteristics, but also remove the noise points.