基于小波技术定量反演多种果树叶片叶绿素含量的研究  

Quantitative Retrieval of Chlorophyll Content in Leaves of Various Fruits and Leaves Based on Wavelet Technology

作　　者：王延仓 石继香 张亮[1] 顾晓鹤[6] 王晶 林家禄 WANG Yancang;SHI Jixiang;ZHANG Liang;GU Xiaohe;WANG Jing;LIN Jialu(North China Institute of Aerospace Engineering,Langfang,Hebei 065000;Shijiazhuang Tiedao University,Shijiazhuang,Hebei 050043;National&Regional Joint Engineering Research Center for Aerospace Remote Sensing Application Technology,Langfang,Hebei 065000;Aerospace Remote Sensing Information Processing and Application Collaborative Innovation Center of Hebei Province,Langfang,Hebei 065000;Shenzhen Intergrated Geotechnical Investigation&Surveying Co.Ltd.,Shenzhen,Guangdong 518109;National Engineering Research Center for Information Technology in Agriculture,Beijing 100097)

机构地区：[1]北华航天工业学院,河北廊坊065000 [2]石家庄铁道大学,河北石家庄050043 [3]河北省航天遥感信息处理与应用协同创新中心,河北廊坊065000 [4]航天遥感信息应用技术国家地方联合工程研究中心,河北廊坊065000 [5]深圳市建设综合勘察设计院有限公司,广东深圳518109 [6]国家农业信息化工程技术研究中心,北京100097

出　　处：《北方园艺》2022年第13期8-15,共8页Northern Horticulture

基　　金：国家重点研发计划资助项目(2019YFE0127300);北京市农林科学院科技创新能力建设专项资助项目(KJCX20170705);国家自然科学基金资助项目(41401419);河北省教育厅青年基金资助项目(QN2019213,ZD2019138)。

摘　　要：以“红叶碧桃”“大金星山楂”“红柿”的果树冠层叶片为试材,通过野外实地试验获取叶片高光谱数据与相应SPAD数据;采用传统数学变换、小波变换处理分析光谱数据,并利用偏最小二乘算法构建叶绿素含量估测模型,研究精准检测多种果树叶绿素含量的方法,分析小波技术分离光谱信息的规律,以期为多种果树的同步精细管理提供参考依据。结果表明:1)小波技术可将不同强度的吸收特征进行二次分配,光谱吸收特征逐步从低频信息转移至高频信息内,且高频信息内涵的吸收特征强度随分解尺度的增加而逐渐增加;2)与原始光谱相比,数学变换、小波技术均能明显提升光谱数据对叶绿素含量的敏感性,二者与叶绿素含量的R^(2)最高可达0.879(位于H_(8)的479 nm处);3)在基于光谱构建的叶绿素含量估测模型中,以小波技术H构建的模型精度最高,为最优模型,其验证样本的R^(2)=0.941,RMSE=0.227,RPD=4.019,这表明利用光谱技术开展多种果树叶片叶绿素含量的精准检测可行且精度较高。Taking the canopy leaves of ‘Red Leaf Green Peach’‘Big Venus Hawthorn’ and ‘Red Persimmon Tree’ as experimental materials, the hyperspectral data and corresponding SPAD data were obtained by field experiment.The traditional mathematical transform and wavelet transform were used to process and analyze the spectral data, and the partial least square algorithm was used to construct the estimation model of chlorophyll content, the method of accurately detecting the chlorophyll content of many kinds of fruit trees was studied, the law of the separation of spectral information by wavelet technology was analyzed, in order to provide reference for the fruit tree management.The results showed that, 1) wavelet technology could redistribute the absorption characteristics of different intensities, and the spectral absorption characteristics were gradually transferred from low-frequency information to high-frequency information, and the intensity of the absorption characteristics of high-frequency information content increases with the increase of decomposition scale.2) Compared with the original spectrum, mathematical transformation wavelet technology could significantly enhance the sensitivity of spectral data to chlorophyll content, and the maximum R^(2) of the two and chlorophyll content could reach 0.879(located at 479 nm of H_(8)).3) In the chlorophyll content estimation model based on spectral construction, the model constructed by wavelet technology Hhad the highest accuracy and was the optimal model, and its validation sample R^(2)=0.941,RMSE=0.227,RPD=4.019,which indicates that the accurate detection of chlorophyll content in various fruit leaves by spectral technology was feasible and accurate.

关 键 词：高光谱 小波技术 数学变换 果树 

分 类 号：S628[农业科学—园艺学]