基于多特征自适应的单光子点云去噪算法

Single photon point cloud denoising algorithm based on multi-features adaptive

新型星载光子计数雷达可获取地面及地面目标的高精度三维信息,但是其测量精度受噪声影响较大。针对在背景噪声不一致及坡度较大区域自动化提取单光子激光数据信号较为困难的难题,文中提出基于多特征自适应的单光子点云去噪算法,有别于传统圆形或椭圆形滤波核,选择更加符合单光子点云数据特征的平行四边形滤波核,分别通过坡度、空间密度、噪声率等多特征自适应识别信号。选择位于青藏高原冰川区域坡度较大且地形破碎的ICESat-2单光子点云数据,开展点云去噪试验和验证,通过与ATL03、ATL08官方去噪结果对比,文中算法在背景噪声水平不一致和大坡度区域具有更优的性能。

The new spaceborne photon counting radar can acquire high-precision three-dimensional information of ground and ground targets, but its measurement accuracy is greatly affected by noise. Aiming at the difficulty of signal extraction of single-photon laser data in areas with inconsistent background noise and large slope area,this paper proposed a single photon point cloud denoising algorithm based on multi-feature adaptive. It was different from the traditional circular or elliptical filtering kernel, and used the parallelogram filtering kernel which was more in line with the characteristics of single photon point cloud data, and signals were adaptively identified by slope, spatial density and noise rate. The ICESat-2 single photon point cloud data located in the glacier area of Qinghai-Tibet Plateau was selected to carry out the point cloud denoising test and verification, and the study area had a large slope and broken terrain. Compared with the official denoising results of ATL03 and ATL08, the proposed algorithm has better performance in areas with inconsistent background noise level and large slope area.