中文摘要：

高光谱遥感技术的出现为森林树种精细识别,特别是针叶林类型识别提供了新的解决思路。通过ASD FieldSpec 3光谱辐射仪对滇西北高寒山区的典型树种云冷杉叶片光谱数据进行测量,并利用光谱微分法对云冷杉原始光谱数据进行处理,分析不同原始光谱、光谱一阶微分、光谱二阶微分曲线图,从中选择能有效区分云冷杉的特征波段,利用欧氏距离对不同方法所选择的波段进行区分度检验。结果表明,能够对云冷杉进行精细识别的特征波段大部分位于近红外波段,原始光谱差异最大波段的位于1 340-1 349nm,一阶微分和二阶微分差异最大波段都位于990-999nm。研究结果将为森林树种精细识别,特别是为针叶林类型精细识别领域提供可靠经验与技术支撑。

英文摘要：

The emergence of the hyperspectral remote sensing technique provides a new solution for precise recognition of forest species,especially the recognition of coniferous forest type. Based on the ASD Field- Spec3 spectral radiometer,the spectroscopic data of blades of spruce-fir forests typically occurring in the al- pine mountainous areas in Northwest Yunnan were measured. The spectrum differential method was adopt- ed to process the original spectroscopic data of the spruce-fir forests. Different original spectrum, spectrum first-order differential and spectrum second-order differential curves were analyzed. Characteristic wave- bands which could efficiently distinguish spruce-fir forests were selected and the Euclidean distance was a- dopted to conduct discrimination test of wavebands selected by different methods. The results suggested that wavebands which could efficiently discriminate features of the forest species were concentrated in the near-infrared region,and that the waveband with the maximum original spectral differences was located in 1 340-1 349 nm. The waveband of the maximum first-order and second-order differential differences was located in 990-999 nm. The results could provide reliable experiences and technical support for precise rec- ognition of the coniferous forest types.