【作者】 乔欣； 马旭；

【Author】 Xin Qiao, Xu Ma (College of Biological and Agricultural Engineering, Jilin University/ Key Laboratory of Terrain-Machine Bionics Engineering, Ministry of Education, Changchun 130025 Jilin China)

【机构】 吉林大学生物与农业工程学院/地面机械仿生技术教育部重点实验室；

【摘要】 简捷、准确、非破坏性的农作物叶绿素和氮素含量诊断在数字农业中具有重要的理论研究价值和实际应用意义。本文以大豆为研究对象,研究其冠层光谱特性,分析大豆冠层光谱反射率与叶片叶绿素含量之间的相关性变化。结果发现:(1)大豆具有绿色植物典型的光谱曲线反射特征;(2)绿光波段540nm、红光波段660nm和近红外波段950nm处光谱反射率与叶绿素指标具有较好的相关性。在此基础上,本研究对比了CARI,SAVI,PVI,NDVI和NIR/G五种光谱植被指数于叶绿素含量的预测模型,筛选出相关性最高的土壤调整植被指数SAVI,并建立最佳估算模型。试验表明,反演模型的相对误差在20%范围内;接着利用叶绿素与N素含量之间存在的相关关系,建立了氮素含量的SAVI反演方程,这将为农田施肥作业提供指导。更多还原

【Abstract】 There are important academic value and actual application significance of Nutrition diagnosis on crop chlorophyll’s content (CHL.C) and nitrogen ’s content (N.C) with simple, exact and nondestructive methods in digital agriculture. First, the reflected spectrum characteristics of soybean’s canopy were taken measure in this paper. Then, the correlative properties between reflected spectrum and CHL.C were analyzed. The conclusions are: 1) the soybean has typical reflected spectral characteristics about green plant. 2) There was a better pertinence between CHL.C and reflected spectrum of 540nm in green, 660nm in red and 950nm in NIR. Five predicted models about CHL.C were established by spectral vegetation indexes of CARI, SAVI, RVI, NDVI, and NIR/G respectively. We got that a predicted model of vegetation Index-SAVI was optimizing. The validation testing showed the relative error of the predicted models is less than 20%. At the same time, we set up the SAVI predicted equation of N.C based on pertinence of N.C and CHL.C. It can be directed fertilizer operation of farm field.更多还原