研究目的
Investigating the feasibility of using laser-induced breakdown spectroscopy (LIBS), near-infrared spectroscopy (NIR), and their joint spectra (LIBS–NIR) for the identification of citrus huanglongbing (HLB).
研究成果
The joint spectra of LIBS and NIR combined with MLP-PCA provided the highest diagnostic accuracy for identifying citrus HLB, demonstrating the potential of optical sensors in the rapid and green detection of citrus diseases.
研究不足
The study is limited by the complexity of citrus matrix effects and the influence of experimental instruments fluctuations. The accuracy of identification could be affected by the variety of citrus, growing environment, and planting means.
1：Experimental Design and Method Selection:
The study employed LIBS and NIR spectroscopy to analyze healthy and HLB-infected citrus leaves. Principal component analysis (PCA) was used to extract characteristic vectors from the spectra. Discriminating analysis (DA) and multi-layer perception (MLP) were combined with PCA to build discrimination models.
2：Sample Selection and Data Sources:
Healthy and HLB-infected fresh leaves were collected from navel orange orchards in Ganzhou of Jiangxi Province, China. A total of 80 leaves (40 healthy and 40 HLB-infected) were selected for analysis.
3：List of Experimental Equipment and Materials:
A Q-switched Nd: YAG pulsed laser (Beamtech, Vlite 200, China) for LIBS, a Fourier transform near-infrared spectrometer (Antaris II, Thermo Scientific, USA) for NIR, and an atomic absorption spectroscopy (AAS) for mineral content analysis.
4：Experimental Procedures and Operational Workflow:
Leaves were cleaned and dried naturally. LIBS and NIR spectra were collected, and mineral content was measured by AAS. Data were analyzed using PCA, DA, and MLP.
5：Data Analysis Methods:
PCA was used for dimensionality reduction. DA and MLP were employed for building classification models to distinguish between healthy and HLB-infected leaves.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
