论文部分内容阅读
辐射归一化旨在减小不同时相遥感影像间因获取条件不一致而导致的非地表辐射变化的差异,是土地覆盖变化监测的重要前提条件。本文根据高光谱图像上同类地物的谱形及数值的相似性,利用光谱角距离(SAD)和欧氏距离(ED)双重判定选取不变特征点,提出了一种基于光谱角—欧氏距离的辐射归一化方法。在评价指标中除了常用的均方根误差和相对偏差,更增加了高光谱特色的衡量光谱保真性指标:皮尔森系数、光谱扭曲程度。利用高光谱遥感CHRIS图像对本文提出方法进行验证,并与基于多元变化检测(MAD)的辐射归一化方法比较。结果表明,本文方法不仅在辐射特性上优于基于多元变化检测(MAD)的方法,而且具有保持光谱特性的优势,具有较好的应用前景。
The aim of radiation normalization is to reduce the difference of non-surface radiation change caused by the inconsistent acquisition conditions of remote sensing images in different temporal phases, which is an important prerequisite for land cover change monitoring. Based on the similarity of spectral shape and numerical value of the similar ground objects in hyperspectral images, this paper selects invariant feature points by the dual judgment of spectral angle (SAD) and Euclidean distance (ED) Distance of radiation normalization method. In addition to the commonly used root mean square error and relative deviation in the evaluation index, the hyperspectral characteristics of the measured spectral fidelity indicators: Pearson coefficient, the degree of spectral distortion. The hyperspectral remote sensing CHRIS image was used to verify the proposed method and compared with the normalized radiation method based on multivariate change detection (MAD). The results show that the proposed method is not only better than the MAD method in terms of radiation characteristics, but also has the advantage of maintaining the spectral characteristics and has good application prospects.