引用本文:
【打印本页】   【下载PDF全文】   View/Add Comment  Download reader   Close
←前一篇|后一篇→ 过刊浏览    高级检索
本文二维码信息
码上扫一扫!
基于连续统去除和偏最小二乘回归的油菜SPAD高光谱估算
郑 煜1, 常庆瑞1, 王婷婷,等1
西北农林科技大学 资源环境学院
摘要:
【目的】探讨油菜叶绿素含量的高光谱估算方法,为实现油菜叶片叶绿素含量的高效、无损、大面积监测提供理论依据。【方法】以陕西省关中地区油菜叶片为研究对象,分别测定苗期、蕾薹期、开花期及角果期的叶片高光谱数据和SPAD值,提取各生育期连续统去除光谱和7类光谱吸收特征参数,分析原始光谱、连续统去除光谱、光谱吸收特征参数与SPAD值之间的相关关系,构建基于原始光谱特征波段、连续统去除光谱特征波段、光谱吸收特征参数的SPAD估算模型,并对模型精度进行验证。【结果】在可见光范围,光谱反射率由蕾薹期、开花期、苗期到角果期依次递增,最大吸收深度和吸收谷面积逐渐增大。利用连续统去除光谱特征波段与吸收特征参数,分别建立的油菜各生育期叶片SPAD估算模型均优于原始光谱。运用连续统去除光谱特征波段结合最优吸收特征参数构建的偏最小二乘回归估算模型,是进行油菜叶片SPAD估算的最优模型。【结论】连续统去除法对不同生育期油菜叶片叶绿素相对含量具有较好的预测能力,是估算油菜叶片SPAD值的一种实时高效方法。
关键词:  连续统去除法  油菜  叶绿素测定  SPAD估算
DOI:
分类号:
基金项目:国家“863”高技术研究与发展计划项目(2013AA102401)
Hyperspectral estimation of SPAD value in oilseed rape based on-continuum removal and partial least squares regression
ZHENG Yu,CHANG Qingrui,WANG Tingting,et al
Abstract:
【Objective】The hyperspectral estimation method of chlorophyll content in oilseed rape was investigated to provide basis for efficient,non-destructive and large area monitoring of chlorophyll content in rape leaves.【Method】Taking rape leaves in Guanzhong,Shaanxi as the research object,the hyperspectral data and SPAD values of leaves at seedling stage,bud stage,flowering stage and pod stages were measured and the continuum removal spectral and 7 spectral absorption characteristic parameters were extracted.The correlations between original spectrum,continuum removal spectrum,spectral absorption characteristic parameters and SPAD were analyzed.SPAD estimation models based on original spectral characteristic band,continuum removal spectral characteristic band and spectral absorption characteristic parameter were constructed and verified.【Result】In the visible light range,the spectral reflectance increased from bud stage,flowering stage,seedling stage to pod stage,and the maximum absorption depth and area of absorption valley increased gradually.The estimation models for each growth period using continuum remove spectral characteristic bands and absorption characteristic parameters were better than the original ones.Partial least squares regression (PLS) model was constructed by removing spectral characteristic bands and optimum absorption characteristic parameters.It was the optimum model for estimating SPAD of rape leaves.【Conclusion】Continuous removal method had better prediction ability for chlorophyll content in rape leaves at different growth stages.It is a real-time and efficient method to estimate the SPAD value of rape leaves.
Key words:  continuum removal method  oilseed rape  chlorophyll determination  SPAD estimation

You are the NO.93582078
Copyright©2009:Editorial Department of Journal of Northwest A&F University (Natural Science Edition)
Designed by Beijing E-Tiller Co.,Ltd