高光谱数据湿地植被类型信息提取

Extraction of wetland vegetation information using hyperspectral image data

以美国Sacramento-San Joaquin三角洲为研究区,利用高光谱和高空间分辨率遥感影像HyMap数据,在光谱特征分析和实测数据的基础上,构造特征指数,建立决策树分类模型对湿地植被进行分类。研究结果表明,湿地植被在近红外波段（0.75-1.3μm）上有明显的光谱特征差异,根据这些差异,可以构造合适的特征指数,实现湿地植被在物种水平上的识别。

In this paper, we mapped wetland vegetation with 3 m spatial resolution, 126-band HyMap image data in California＇s Sacramento-San Joaquin delta. Specific vegetation indices were constructed and a decision tree model was used to identify wetland vegetation based on spectral analysis and field investigation. The result showed that we could construct suitable specific vegetation indices, according to the spectral difference obvious at near infrared bands （0.75- 1.3 μm）, which could be effective in distinguishing wetland vegetation and allowing for species-level detection necessary to map invasive species.