论文部分内容阅读
冰川是自然界重要而且极具潜力的淡水资源,对区域生态平衡与稳定起着重要作用。本研究获取了2011年8月青藏高原“中习一号”冰川的机载高光谱数据。首先对该数据进行预处理,包括辐射定标、大气校正和几何校正;然后分别采用主成分分析(PCA)和最小噪音变换(MNF)两种方法进行高光谱数据降维处理。在此基础上,分别利用马氏距离法、最大似然法、最小距离法、波谱角法、二进制编码法和光谱信息散度6种图像分类方法进行冰川分类,并对分类结果进行比较,确定本研究区的最优数据降维方法和分类方法;最后将高光谱数据分类结果与环境小卫星多光谱数据分类结果进行了比较分析。研究结果表明:高光谱数据经PCA变换后的分类效果优于MNF变换后的分类结果;经PCA变换后的数据采用马氏距离法、最大似然法、最小距离法分类效果较好;MNF变换后采用波谱角法和光谱信息散度效果较好。
Glacier is an important and potentially freshwater resource in nature and plays an important role in regional ecological balance and stability. In this study, we acquired the airborne hyperspectral data of the Qinghai-Tibet Plateau in August 2011. Firstly, the data is preprocessed, including radiation calibration, atmospheric correction and geometric correction. Then the principal component analysis (PCA) and minimum noise transform (MNF) are respectively used to reduce the dimension of hyperspectral data. On this basis, the glaciers are classified by using six kinds of image classification methods, such as Mahalanobis distance method, maximum likelihood method, minimum distance method, spectral angle method, binary coding method and spectral information divergence, and the classification results are compared and determined The best data reduction method and classification method in this study area; and finally, the classification results of hyperspectral data and the comparative analysis of environmental multi-spectral data of small satellites were analyzed. The results show that the classification result of hyperspectral data after PCA transform is superior to the classification result after MNF transform. The data transformed by PCA is better than Malay distance method, the maximum likelihood method and the minimum distance method. The MNF transform Spectral angle and spectral information after the divergence effect is better.