论文部分内容阅读
利用黑龙江省伊春市带领区凉水国家级自然保护区机载高光谱数据,提取了红边面积、三角形植被指数、归一化植被指数等15个光谱参数,结合坡度、坡向、海拔、郁闭度和植被总盖度5个地理参数,并利用叶绿素计SPAD-502对研究区植被冠层叶绿素相对含量进行同步测量,分析了叶片光谱反射率、反射率的一阶导数及其他变形分别与SPAD值的相关性,采用基于核变换的偏最小二乘原理建立了叶绿素相对含量的估测模型,用该模型对研究区植被冠层叶绿素相对含量进行定量估算.结果表明:当分段数为3、提取的主成分数为10时,所建模型的效果较好,模型决定系数达到0.855,平均绝对百分误差为9.6%,预测精度为89.7%.
Fifteen spectral parameters such as red edge area, triangular vegetation index and normalized vegetation index were extracted by using airborne hyperspectral data from Liangshui National Nature Reserve, a leading area of Yichun City, Heilongjiang Province. Combining with slope, aspect, altitude, Closed degree and total vegetation coverage. The chlorophyll meter SPAD-502 was used to measure the relative content of chlorophyll in the canopy of the study area. The first derivative and other deformations of spectral reflectance and reflectance were analyzed separately SPAD value, the estimation model of relative content of chlorophyll was established based on the principle of partial least squares with kernel transform, and the relative content of chlorophyll in the vegetation canopy was quantitatively estimated by this model.The results showed that when the number of segmentation was 3, When the number of principal components extracted is 10, the effect of the proposed model is better. The model determination coefficient reaches 0.855, the average absolute percentage error is 9.6% and the prediction accuracy is 89.7%.