利用机载点云检核ICESat-2/ATLAS激光测高数据精度

Verification of ICESat-2/ATLAS Laser Altimetry Data Accuracy Using Airborne Point Cloud

美国第2代激光测高卫星ICESat-2已经公开发布数据,能够提供剖面状密集点云,但其平面/高程精度还未得到充分验证。全球开源DEM或卫星光学立体影像生成的DEM精度远低于星载激光测高数据,无法实现激光精度检核。本文使用机载激光雷达/光学立体影像数据开展ICESat-2卫星ATLAS精度验证,通过单光子点云去噪、剖面地形恢复和数据内插等手段,对国内舟山、上海和汉中等不同地形条件下的星载光子计数激光雷达数据进行对比分析。实验结果表明,6组ATLAS点云高程的均方根误差分别为0.43、0.65、0.72、0.53、0.45和0.77 m,平均值达到0.59 m,具备较好的高程精度。

The second generation of the US laser altimetry satellite ICESat-2 has released data to provide a dense point cloud,but its plane/elevation accuracy has not been fully verified.The accuracy of DEM produced by traditional satellite optical stereo image or the global open DEM is far lower than that of satellite laser altimetry data,so it is impossible to achieve accuracy verification.The airborne lidar/optical stereo image is used to verify the accuracy of ATLAS data of ICESat-2 satellite in the paper.By means of single photon denoising,profile terrain recovery and data interpolation,photon counting lidar point data(ATLAS 03)under different terrain conditions in Zhoushan,Shanghai and Hanzhong are compared and analyzed.The results show that the root mean square errors of ATLAS point cloud elevation of 6 groups data are 0.43,0.65,0.72,0.53,0.45 and 0.77 m respectively,with a mean value of 0.59 m,achieving good elevation accuracy.