论文部分内容阅读
为了解决混合像元问题,发展了一门新的技术——光谱分离技术,线形混合模型是目前应用最成功的方法,但是当应用到月球风化层遥感时,其存在两个内在的缺陷一是由于充分混合引起的非线性,二是纯光谱的不“纯”性.对于第一个缺陷,从Hapke理论出发,充分考虑了数据的特点,大大简化多重散射模型来提高充分混合的线形性,同时降低了原模型计算的复杂度;对于后一个问题,提出用修正线形混合模型来弥补“纯”光谱中的误差.该方法通过Relab数据库中的模拟月壤的光谱数据进行了验证,试验结果表明,这种方法具有很好的性能.
In order to solve the mixed pixel problem, a new technology, spectral separation technology, is developed. The linear mixed model is the most successful method applied at present. However, there are two inherent defects when applied to the remote sensing of the lunar weathering layer. Due to the non-linearity caused by the thorough mixing, the second one is not the “pure” nature of the pure spectrum. For the first defect, based on the Hapke theory, the characteristics of the data are sufficiently considered, the multiple scattering model is greatly simplified to improve the mixed linearity, At the same time, the computational complexity of the original model is reduced. For the latter problem, a modified linear mixed model is proposed to make up for the error in the “pure” spectrum. The method is verified by the spectral data of simulated lunar soil in the Relab database. Show that this method has good performance.