摘要
针对难以完整地提取密度不均匀的浅海ICESat-2光子数据,该文以我国南海的全富岛和珊瑚岛为例,提出浅海ICESat-2光子数据的投影去噪方法。首先根据浅海水下地形对光子数据分组,然后将各组数据投影在等高线附近,并通过高斯拟合去除噪声。最后将光子投影去噪方法与Hsu提出的去噪方法进行对比,并结合Sentinel-2遥感影像数据开展主被动水深反演进行验证。实验表明,光子投影去噪能够准确地提取出浅海水下信号光子数据,时间效率上优于Hsu的去噪方法。利用该去噪方法得到的水深控制点建立的水深反演模型反演误差优于0.9 m,相对精度小于9%。光子投影去噪方法能够为浅海ICESat-2光子数据去噪工作提供参考。
In view of the difficulty in extracting ICESat-2 photon data with uneven density in shallow sea,a projection denoising method for shallow sea ICESat-2 photon data is proposed in this paper,with the full-rich island and coral island in the South China Sea as an exemplar.Firstly,the photon data are grouped according to the shallow underwater topography,then the data are projected near the contours,and the noise is removed by Gaussian fitting.Finally,the photon projection denoising method is compared with the denoising method proposed by Hsu,and verified by active/passive water depth inversion combined with Sentinel-2 remote sensing image data.The experimental results show that photon projection denoising can accurately extract the photon data of underwater signal in shallow sea,which is better than Hsu's denoising method in time efficiency.The inversion error of the water depth inversion model established by the water depth control points obtained by the denoising method is better than 0.9 m,and the relative accuracy is less than 9%.The photon projection denoising method can provide reference for ICESat-2 photon data denoising in shallow water.
作者
白龙
黄文骞
BAI Long;HUANG Wenqian
出处
《科技创新与应用》
2022年第33期1-5,共5页
Technology Innovation and Application
基金
国家自然科学基金(41871295)。
