减小由光谱测量数据误差造成的薄膜光学参数反演不确定度的方法

基于对光谱测量数据中系统误差和随机误差的不同特性分析,提出了两种方法来分别减小其对薄膜光学参数反演测量精度的影响。方法一是利用薄膜光谱系数关于薄膜光学参数的一阶偏导数信息筛选数据点,选取偏导数符号相反的波段或入射角区域作为最佳测量数据点采集区域,以最小化系统误差引起的薄膜光学参数相对真实值的偏差大小。方法二是在实测光谱数据中注入独立同分布的随机噪声,利用基于偏差函数最小化方法的混合优化算法多次进行反演计算,将每次反演得到的薄膜光学参数估计值的统计均值作为薄膜真实光学参数的估计值,以减小甚至消除随机误差引起的薄膜光学参数不确定度。这两种方法不仅在提高薄膜光学参数测量精度上具有明显的实用价值,而且物理意义明确,可操作性强,不局限于特定波段和特定材料的薄膜光学参数测量,具有良好的普适性,可望在薄膜光学参数的准确测量技术中发挥重要作用。

Some Peculiarities of the Preprocessing of Spectral Data and Images

Remotely sensed spectral data and images are acquired under significant additional effects accompanying their major formation process, which greatly determine measurement accuracy. In order to be used in subsequent quantitative analysis and assessment, this data should be subject to preliminary processing aiming to improve its accuracy and credibility. The paper considers some major problems related with preliminary processing of remotely sensed spectral data and images. The major factors are analyzed, which affect the occurrence of data noise or uncertainties and the methods for reduction or removal thereof. Assessment is made of the extent to which available equipment and technologies may help reduce measurement errors.