Review of Noise Filtering Algorithm For Photon Data

As a continuation of Ice, Cloud, and Land Elevation Satellite-1 (ICESat-1)/Geoscience Laser Altimeter System (GLAS), the Ice, Cloud, and Land Elevation Satellite-2 (ICESat-2) , which is equipped with the Advanced Topographic Laser Altimeter (ATLAS) system, was successfully launched in 2018. Since ICESat-1/GLAS has facilitated scientific results in the field of forest structure parameter estimation, how to use the ICESat-2/ATLAS photon cloud data to estimate forest structure parameters has become a hotspot in the field of spaceborne photon data application. However, due to the weak photon characteristics of the ICESat-2/ATLAS system, the system is extremely susceptible to noise, which poses a challenge for its subsequent accurate estimation of forest structural parameters. Aiming to filter out the noise photons, the paper introduces the advantages of the spaceborne lidar system ICESat-2/ATLAS than ICESat-1/GLAS. The paper summarizes the research of the simulated photon-counting lidar (PCL) noise filtering algorithm and noise filtering on spaceborne.


Forest structural parameters have been listed as one of the important indicators for monitoring forest carbon by the International Union of Forest Research
Organizations (IUFRO). How to quickly and accurately quantificate the forest structure parameters of the study area is great significance for monitoring forest carbon (Jandl, 2019 andGolshani, 2019). In the process of forest monitoring, forest structural parameters (such as canopy height, et al.) directly reflect the growth of forests, and have become one of the most commonly used bases in forest ecological research (Wu, 2019). Traditional forest monitoring methods are mostly acquired by manual field measurement. Although the measurement is accuracy, it has a long working cycle and low efficiency and not applicable complete large-scale, multi-scale for forest structure parameter monitoring (Piermattei, 2019) .
In recent years, with the development of remote sensing technology, remote sensing technology has become a key technology for monitoring carbon, and has been identified as a monitoring method that complies with the "Kyoto Protocol" (Lefsky, 2002).

2019) by cloud and biomass saturation, spaceborne
Light Detection And Ranging (LiDAR) technology overcomes the above problem from the perspective of data source, and has been widely used with its efficient and accurate ranging capability. In order to acquire large-area, multi-scale, multi-temporal forest monitoring data, studies used the spaceborne lidar data to retrieve forest structural parameters. So far, two spaceborne lidar could provide scientific data, include Ice, Cloud, and Land Elevation Satellite (ICESat-1) launched by the NASA in 2003, , Hu, 2019and Lefsky, 2010 (Sawruk, 2018 andMarkus, 2017). However, because the ICESat-2 system has the characteristic of photon-counting, the signal photons emitted and received are weak signal photons, which are greatly affected by background noise (solar background noise, system dark noise, et al.). Therefore, how to effectively filiter noise photon is the hotspot in photon data processing and application (Li, 2018, Huang, 2019and Moussavi, 2014.
In order to advance the progress of the spaceborne photon noise filitering algorithm, this paper will introduce the advantages of the spaceborne lidar system ICESat-2 and its relative ICESat-1/GLAS, and The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-3/W10, 2020 International Conference on Geomatics in the Big Data Era ( which is more susceptible to information such as terrain slope (Yang,2019, Moussavi,2014and Yang,2019. Unlike the waveform technology used in the ICESat-1/GLAS system, the ICESat-2/ATLAS system will use photon-counting technology. The GLAS laser has high pulse energy, low frequency, and measured target information by waveform. The ICESat-2/ATLAS has low pulse energy and high frequency. The recorded target information is determined by recording the returned photon time tag information. The ICESat-2/ATLAS system has three advantages over the ICESat-1/GLAS system. Firstly, its photon density is denser than the ICESat-1/GLAS system (ATLAS is 0.7m along the foot footprint, GLAS is 170m along the foot footprint), could acquire continuous information, and the information of the measured target can be more accurately described, which is significance for improving the extraction of forest structure parameters. Secondly, the ICESat-2/ ATLAS system is a multi-beam transmission method, and could acquire more observation information under the same orbit. Thirdly, the ICESat-2/ATLAS system has a smaller footprint diameter (17m for ATLAS and 70m for GLAS), which reduces the impact of slope on spot information acquisition.

Filitering Algorithm
In order to simulate the ICESat-2/ATLAS spaceborne photon data noise filitering algorithm, the noise filtering method based on SPL data as experimental data, and determined the optimal voxel size for the optimal filter by testing different voxel sizes. It is finally determined that Garrett algorithm is the optimal algorithm for the experimental data, and 3m×3m×0.2m is the optimal voxel size of the test data.
This method effectively filters the noise photon of the smooth surface while maintaining the spatial integrity of the SPL data, but the algorithm has filitered less photon in the research area with forest cover on the surface, and the SPL data is high density photon cloud data. There is a significant difference between the SPL data and the ATLAS photon data, so the voxel-based spatial filtering algorithm is not necessarily applicable in the ATLAS photon data. Due to the ICESat-2/ATLAS photon density characteristics, the photon density in the horizontal direction is higher than the vertical direction.

CONCLUSIONS
According to the research, it is recommended to pay attention to the following problems in the future study: 1) The photon noise filtering algorithm has high precision on smooth surfaces, while in the forest coverage area, due to noise photons distributed over the canopy, inside the canopy and below the ground, which seriously affects the accuracy of the noise filtering algorithm. 2) The above research only selects one type of airborne route. The simulated PCL data is not analysis noise filtering by laser pointing and strong and weak type, lacking comprehensive analysis and evaluation of noise filtering effect. 3) Most algorithms need to manually adjust the parameter of the noise filtering algorithm. It also limits the adaptability of the algorithm. Therefore, how to improve the noise filtering effect of the spaceborne photon data in the forest coverage area by an adaptive way is the technical difficulty in photon data processing.