研究目的
To develop a new spectral unmixing framework that incorporates LiDAR data into the weighting of the spatial regularization and to conduct a comprehensive comparison of the weighting functions derived from LiDAR data or from its combination with another guidance map.
研究成果
The incorporation of LiDAR data into the spatial regularization of spectral unmixing significantly improves the accuracy of abundance estimates, demonstrating robustness and effectiveness across different simulated scenarios.
研究不足
The study is limited to simulated data, and the effectiveness of LiDAR data in spatial regularization may vary with the accuracy of DSM and the spectral similarity of materials in real-world scenarios.
1：Experimental Design and Method Selection:
The study employs a linear mixture model (LMM) for spectral unmixing, incorporating spatial regularization informed by LiDAR data.
2：Sample Selection and Data Sources:
Two simulated datasets (SIM1 and SIM2) are used, one with synthetic DSM and hyperspectral data, and another with real DSM and synthetic hyperspectral data.
3：List of Experimental Equipment and Materials:
Hyperspectral and LiDAR data are the primary materials.
4：Experimental Procedures and Operational Workflow:
The method involves adjusting weights in the spatial regularization term based on LiDAR-derived DSM and comparing performance against methods using hyperspectral image-derived spatial information.
5：Data Analysis Methods:
Root mean square error (RMSE) of abundance estimates is used for quantitative validation.
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
