病害胁迫下冬小麦冠层叶片色素含量高光谱遥感估测研究

Study on Hyperspectra Estimation of Pigment Contents in Canopy Leaves of Winter Wheat Under Disease Stress

通过人工田间诱发小麦条锈病,在不同生育期测定染病冬小麦冠层光谱和相应叶片的色素含量。把冠层光谱数据、一阶微分数据与相应的叶片色素含量数据分别进行相关分析,采用单变量线性和非线性回归技术,选取部分样本建立小麦的色素含量估测模型,并利用其余的样本对模型进行检验,结果表明绿边内一阶微分总和(SDg)与红边内一阶微分总和(SDr)的归一化值为变量的线性模型是估测色素含量的最佳模型,其估测叶绿素a,叶绿素b和胡萝卜素含量的相对误差分别为17.0%,16.3%和12.4%。该研究表明可用高光谱信息估测冠层叶片色素含量,且估测精度较高。文章的研究结果对利用高光谱遥感监测农作物长势以及病害影响都具有实际应用价值。

The canopy reflectance of winter wheat infected with stripe rust was measured in the field through artificial inoculation, and the pigment contents of the wheat leaves were determined indoor. The correlation between pigment contents and canopy hyperspectra data and the first derivative data of the disease wheat were analyzed respectively. Using linear and non-linear regression methods, and choosing a part of samples, the estimation models about pigment contents of disease wheat were built. Through the test of the other part samples, the result shows that the model containing the normalized value of the sum of first derivative within green edge （SDg） and the sum of first derivative within red edge （SD,） is the best one. The model was used to estimate the contents of chlorophyll a and chlorophyll b and carotenoid of the disease wheat, and the relative errors were 17. 0%, 16.3% and 12.4%, respectively. This study shows that canopy hyperspectra data can be used to estimate the pigment contents of crops leaves and the estimation precision is high. This conclusion has great practice and application value to monitor the growing way of and disease influence on crops by using hyperspectral remote sensing.