林区米级空间分辨率机载高光谱影像BRDF校正研究

米级空间分辨率机载高光谱影像，既可提供刻画森林冠层结构的空间信息又具备反映其生理状况的窄波段光谱，在大面积森林资源监测和参数提取中具有重大潜力。但受到太阳-观测几何、地形起伏以及空间尺度的影响，林区米级空间分辨率机载影像会呈现出更为复杂的二向反射分布（BRDF）效应，限制了其大范围的应用。本项目拟结合精细分类数据、LiDAR提取的冠层结构以及冠层光照条件信息，改进适用于米级空间分辨率像元冠层或亚冠层观测尺度的Ross-Li散射核函数。再结合DEM信息发展起伏地形林区机载高光谱影像核驱动BRDF校正方法，分析不同空间分辨率的DEM、DSM对影像BRDF校正结果的影响。最后，拟分别以云南太阳河国家公园和河北塞罕坝林场为试验区开展方法研究，并结合地面调查和无人机多角度观测冠层光谱数据对模型结果进行验证。本研究可以促进大区域森林机载高光谱数据的遥感应用，为我国森林高质量的发展提供科学支撑。

Airborne hyperspectral images with meter-scale spatial resolution can provide the capability to describe the canopy structure and monitor the physiological conditions through some narrow bands spectrum, which has great potential for forest resources surveying over large area. However, the airborne hyperspectral images with meter-scale spatial resolution contain more complex bidirectional reflectance distribution function (BRDF) effect due to the strong influence of viewing-illumination geometry, rugged terrain and fine spatial scale, which clearly limits its large scale use. With the information of fine classification data, LiDAR-derived canopy complexity and incoming irradiance of a canopy, this project intends to improve the Ross-Li kernel function for the canopy or sub-canopy scale within the meter-scale spatial resolution pixel. With the combination of improved Ross-Li kernels and DEM data, a new kernel driven BRDF correction method for airborne hyperspectral images with meter-scale spatial resolution in rugged terrain over forest region will be developed. The sensitivity of BRDF correction to the DEM and DSM spatial scale will be performed. We plan to take the Taiyanghe National Forest Park in Yunnan Province and Saihanba forest farm in Hebei Province as intensive study area to develop algorithms, respectively. These developed BRDF methods will be applied for the airborne images of study area. Then the BRDF corrected results will be compared with the spectral data from ground survey and UAV-based multi-angle hyperspectral sensor. This study can promote the remote sensing application of airborne hyperspectral images over large areas, and provide the scientific support for the high quality development of forestry in China.