中国农业科学 ›› 2013, Vol. 46 ›› Issue (16): 3504-3513.doi: 10.3864/j.issn.0578-1752.2013.16.022

• 研究简报 • 上一篇    下一篇

基于高光谱的苹果盛果期冠层叶绿素含量监测研究

 房贤一, 朱西存, 王凌, 赵庚星   

  1. 山东农业大学资源与环境学院/土肥资源高效利用国家工程实验室,山东泰安 271018
  • 收稿日期:2012-12-04 出版日期:2013-08-15 发布日期:2013-05-28
  • 通讯作者: 通信作者朱西存,E-mail:zxc@sdau.edu.cn
  • 作者简介:房贤一,E-mail:fangzi0425@126.com
  • 基金资助:

    国家自然科学基金(41271369)、山东农业大学青年科技创新基金(23731)、山东省自然科学基金(ZR2012DM007)、山东农业大学博士后基金(89841)、中国博士后基金(20110491616)

Hyperspectral Monitoring of the Canopy Chlorophyll Content at Apple Tree Prosperous Fruit Stage

 FANG  Xian-Yi, ZHU  Xi-Cun, WANG  Ling, ZHAO  Geng-Xing   

  1. College of Resources and Environment, Shandong Agricultural University/National Engineering Laboratory for Efficient Utilization of Soil and Fertilizer, Tai’an 271018, Shandong
  • Received:2012-12-04 Online:2013-08-15 Published:2013-05-28

摘要: 【目的】建立苹果冠层叶绿素含量及冠层光谱特征参量间的定量关系模型,以促进高光谱技术在苹果树精准施肥以及快速、无损长势监测中的应用。【方法】以蒙阴县果园的苹果树为试验材料,连续2年分别测定了苹果冠层光谱反射率和冠层叶绿素(Chl(a+b))含量,分析了冠层叶绿素含量与光谱反射率之间的相关关系,并计算了400—1 000 nm任意两波段组合而成的RVI、DVI、NDVI和RDVI,分析了它们与冠层叶绿素含量的关系,以逐步回归分析做比较,建立了苹果冠层叶绿素含量监测模型。【结果】结果表明,以单变量估算叶绿素含量的最佳光谱指数为NDVI(975,742),相关系数为0.5093。利用多元逐步回归建立的苹果冠层叶绿素含量最佳监测模型为Y=-0.56(log1/R)771-0.48(log1/R)1978 +0.20(log1/R)2407 -0.10(log1/R)2440+4.749。【结论】用多元逐步回归方法建立的模型来监测苹果冠层叶绿素含量效果较好,为利用高光谱技术监测苹果生长状况提供了理论依据。

关键词: 苹果冠层 , 高光谱 , 叶绿素含量 , 光谱指数 , 逐步回归

Abstract: 【Objective】 The aims of the study are to promote the application of hyperspectral techniques in the real-time non-destructive detection of the apple precise fertilization and the diagnosis of the growth of the apple tree by building the quantitative relationship models between the apple canopy chlorophyll content and the characteristics of the canopy spectrum . 【Method】 Taking the apple trees in the Mengyin area as the experimental materials, the spectral reflectance and the apple canopy chlorophyll Chl(a+b) content were measured for two years continuously, the correlation coefficient between them was analyzed, and the RVI, DVI, NDVI and RDVI were calculated by combining any two bands from 400 nm to 1 000 nm. The relationship between them was analyzed and stepwise regression method was used to establish the apple canopy chlorophyll content monitoring model. 【Result】 The results showed that the best spectral index to estimate the content of the chlorophyll is NDVI(975,742), the correlation coefficient is 0.5093. The best apple canopy chlorophyll content monitoring model is Y=-0.56(log1/R)771-0.48 (log1/R)1978+0.20(log1/R)2407-0.10(log1/R)2440+4.749. 【Conclusion】 The model established by using stepwise regression can predict the apple canopy chlorophyll content better, thus it has provided a theoretical basis for reflecting apple growth conditions by using hyperspectral technology.

Key words: apple canopy , hyperspectrum , chlorophyll content , spectral indices , stepwise regression