中文摘要：

对不同胁迫条件下棉叶光谱特征进行比较研究,可为棉花胁迫类型识别提供理论依据。以田间获取的9种胁迫棉叶为材料,对9种胁迫叶片进行光谱测试,分析各自光谱特征差异,对其进行定性定量的识别。结果表明：黄萎病、枯萎病、干旱、缺氮、红蜘蛛和角斑病棉叶光谱反射率形状和大小在全波段均发生了较大变化,蚜虫、冻害和盐害均变化较小。9种胁迫棉叶的光谱＂绿峰＂和＂红边＂均具有＂单峰＂现象,其中黄萎病、干旱、缺氮、枯萎病、红蜘蛛和角斑病2个＂单峰＂峰值均减小,＂单峰＂位置均发生＂蓝移＂,而蚜虫、冻害和盐害基本不变。蚜虫棉叶的吸收波段位置、吸收深度、吸收面积略微增加,其他胁迫均减小,蚜虫棉叶吸收宽度略微减小,冻害棉叶吸收宽度基本不变,盐害棉叶吸收宽度先增后减,其他胁迫棉叶的吸收宽度均增加。不同胁迫棉叶的14个光谱特征参数表现出一定的规律,建立相应的判别式可有效对不同胁迫棉叶进行准确识别。因此,利用光谱技术进行棉花叶片不同胁迫诊断是可行的。

英文摘要：

Comparison of spectral characteristics of cotton leaves under different stresses will provide theoretic foundation for recognizing stress type of cotton.Cotton leaves spectrum were obtained from experiment plots and fields under the condition of nine stresses,and their spectra characteristics were analyzed qualitatively and quantitatively.The main results indicated that,in whole spectral bands,the reflectance value and shape of cotton leaf had larger changes under the condtion of Verticillium wilt,Fusarium wilt,drought,nitrogen scarcity,red spider and angular leaf spot,however,minor changes under the condition of aphid,frost injury and salt injury.Cotton leaves of nine stresses had all＂single peak＂phenomenon in＂green peak＂and＂red edge＂,the values of two＂single peak＂all decreased and their position all moved to the blue for the stress of Verticillium wilt,Fusarium wilt,drought,nitrogen scarcity,red spider and angular leaf spot,the values of two＂single peak＂were basic invariability for the stress of aphid,frost injury and salt injury.Absorb band position,absorb depth,and absorb area of aphid-injured leaf had a little increase,which were decrease for others stresses.Absorb width of aphid-injured leaf had a little decrease,frost injury basic invariability,salt injury firstly increases and then decreases,but others stresses all increases.The 14 spectral characteristic parameters were present better change rule to cotton leaves of different stresses.Base on those,the discriminate models were established of nine stresses and can accuracy diagnose them.So,cotton leaves suffered from different stresses could be reorganized by hyper-spectral technology.