基于一种改进的多端元混合像元分解方法的三江源中东部地区植被盖度信息的提取

Extraction of the Vegetation Fraction based on an Improved Multiple Endmember Spectral Mixture Analysis for the Central and Eastern Area of Source Region of Yangtze,Yellow and Lantsang Rivers

三江源地处青海省南部,近年来生态环境不断恶化、草地退化严重。植被盖度是研究草地退化的重要指标之一,如何快速、准确、大面积地获取植被信息对三江源的生态环境变化监测尤为重要。传统的固定端元混合像元分解(TSMA)不适用于地域辽阔、地物类型复杂多样的三江源区,而多端元混合像元分解(MESMA)允许端元的类型和数量随像元的不同而变化,更符合三江源的实际情况。基于已有相关研究提出一种改进的多端元混合像元分解(IMESMA),即加入端元组分合理性最优端元模型判断规则,分两步对影像进行分解并获得像元的植被盖度信息。第一步利用多端元模型探测出含有植被信息和完全不含植被信息的像元,并对非植被像元进行掩膜,以提高分解精度;第二步仅对含有植被信息的像元进行分解,获得最终的三江源中东部植被盖度信息。通过多端元分布图、混淆矩阵和与实测数据的均方根误差(RMSE)对比表明,相比TSMA,IMESMA考虑了同物异谱现象,并且其分解精度高、分解结果可信,更适合三江源区的植被信息提取。本方法也可用于其他复杂环境地表组分信息的提取。

The source region of Three-River,as the largest ecological function district in China,has been deteriorating seriously as a result of human beings＇ activities and the global climate change.The vegetation fraction is a vital indicator for grassland degradation research.It is important to derive the vegetation information quickly,impersonalityly,and dispense with large manpower for field survey.The traditional spectral mixture analysis（TSMA） method is not suitable for the region such as the source region of three rivers that has large area and complex and diverse terrestrial features.However,Multiple endmember spectral mixture analysis is well suited for real world because it allows the number and types of endmembers to vary on a per-pixel bases.In this paper,an improved approach for the extraction of vegetation information for the central and eastern area of the Source Region of Yangtze,Yellow and Lantsang Rivers is suggested.In this approach,the rationality rule of endmember constitute was added to select the optimal endmember model through two steps.MESMA was used to detect the pixels that didn＇t contain vegetation information at all in the first step and these pixels would be masked out then.In the second step,MESMA is used to unmixing vegetation information precisely.The results indicate that the Improved Multiple Endmember Spectral Mixture Analysis（IMESMA） is more effective than TSMA for extraction information of vegetation through the distribution of the multipile endmember models,confusion matrix and the field observational data.The method the paper suggested can also be used to derive terrestrial fraction for other complex region.