研究目的
To study several typical noise filtering methods for photon counting laser altimeter data and discuss their characteristics, with experiments conducted on both ice sheet and vegetation data sets.
研究成果
For complex environment and terrain, the density-dimension method and improved localized statistics-based method are proposed. For ice sheet region data with simple environment, all the methods mentioned are applicable. Future improvements can focus on calculating slope to speed up the calculation speed of the method.
研究不足
Different methods have different effects depending on the environmental conditions. Some methods are not applicable to complex data or when the noise is not uniformly distributed. The parameters used in the methods can greatly influence the results.
1：Experimental Design and Method Selection:
Five denoising methods were compared, including local distance statistics method, histogram based method, modified DBSCAN method, density-dimension method, and improved localized statistics method.
2：Sample Selection and Data Sources:
Two experimental data sets from the simulated ICESat-2 data (MABEL) were used, collected near Northwest Greenland ice sheet and near San Luis Obispo, CA.
3：List of Experimental Equipment and Materials:
MABEL system carried on NASA ER-2 aircraft.
4：Experimental Procedures and Operational Workflow:
Each method was applied to the two data sets, and the results were compared.
5：Data Analysis Methods:
The effectiveness of each method was evaluated based on the denoising results for different environmental conditions.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
