基于统计直方图两步法的星载单光子数据去噪

针对星载单光子激光有效回波信号混杂于噪声中难以区分的问题，提出一种基于统计直方图两步法的星载单光子数据去噪方法。先后采用沿轨小窗口直方图粗去噪与距离平方统计直方图精去噪，实现星载单光子回波数据中的噪声光子有效剔除。利用该方法对美国星载单光子激光雷达先进地形激光测高系统(ATLAS)的强与弱波束、白天和夜间、平地与山地3种典型情况下的回波光子数据进行实验，结合ATLAS官方去噪结果，基于混淆矩阵统计去噪精确度。实验结果表明，强波束数据去噪精确度为98.86%，弱波束数据去噪精确度为96.94%；夜间数据去噪精确度为99.02%，白天数据去噪精度为98.86%；山地数据去噪精确度为96.28%，平地数据去噪精确度为96.94%。说明本文方法适用于常见的以上3种典型情况下的星载单光子数据去噪。

Denoising of spaceborne single-photon data based on two-step method of statistical histogram

It is difficult to distinguish the spaceborne single-photon laser echo signal mixed with the noise. A two-step method for denoising spaceborne single-photon data based on statistical histogram is proposed. In order to eliminate noisy photons in the spaceborne single-photon echo data, a small window histogram method along the track is used for coarse denoising, and then a distance square statistical histogram method is used for fine denoising. The echo photon data of the Advanced Topographic Laser Altimeter System(ATLAS) spaceborne single photon lidar under three typical conditions of strong and weak beam, day and night, flat ground and mountain are selected as experimental data. Combined with the official results of ATLAS and based on the confusion matrix, the de-noising accuracy is calculated. Experimental results show that the denoising accuracy of strong beam data is 98.86%, and that of weak beam data is 96.94%; the denoising accuracy of night data is 99.02%, and that of daytime data is 98.86%; the denoising accuracy of mountain data is 96.28%, and that of flat data is 96.94%. The results show that the proposed method is suitable for spaceborne single photon data denoising under above three typical conditions.