Premium
Discrimination of nitrogen fertilizer levels of tea plant ( Camellia sinensis ) based on hyperspectral imaging
Author(s) -
Wang Yujie,
Hu Xin,
Hou Zhiwei,
Ning Jingming,
Zhang Zhengzhu
Publication year - 2018
Publication title -
journal of the science of food and agriculture
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.782
H-Index - 142
eISSN - 1097-0010
pISSN - 0022-5142
DOI - 10.1002/jsfa.8996
Subject(s) - hyperspectral imaging , camellia sinensis , principal component analysis , support vector machine , imaging spectrometer , artificial intelligence , fertilizer , pattern recognition (psychology) , mathematics , environmental science , biological system , computer science , remote sensing , spectrometer , agronomy , horticulture , biology , geography , physics , quantum mechanics
BACKGROUND Nitrogen (N) fertilizer plays an important role in tea plantation management, with significant impacts on the photosynthetic capacity, productivity and nutrition status of tea plants. The present study aimed to establish a method for the discrimination of N fertilizer levels using hyperspectral imaging technique. RESULTS Spectral data were extracted from the region of interest, followed by the first derivative to reduce background noise. Five optimal wavelengths were selected by principal component analysis. Texture features were extracted from the images at optimal wavelengths by gray‐level gradient co‐occurrence matrix. Support vector machine (SVM) and extreme learning machine were used to build classification models based on spectral data, optimal wavelengths, texture features and data fusion, respectively. The SVM model using fused data gave the best performance with highest correct classification rate of 100% for prediction set. CONCLUSION The overall results indicated that visible and near‐infrared hyperspectral imaging combined with SVM were effective in discriminating N fertilizer levels of tea plants. © 2018 Society of Chemical Industry