基于LIBS和THz谱图结合的油茶隐性病害早期诊断方法研究

With all Camellia oleifera production areas generally affected by diseases, so disease has brought to Camellia oleifera industry an incalculable economic loss,important domestic plant quarantine has become the research hot spot. Looking for fast, real-time, accurate and low cost Camellia oleifera disease diagnosis mechanism and method, realize the early diagnosis and timely treatment, protection of citrus production safety and promote the increase of farmers' income has important value.The project to recessive diseases of Camellia oleifera as the object of study by Laser- induced breakdown spectroscopy combined with terahertz technology carry out recessive diseases of Camellia oleifera of rapid diagnosis mechanism and method of the research. Looking for infected early leaf nutritional elements components of the Laser- induced breakdown spectral changes on a regular same time, three physiological indexes in research early recessive diseases of Camellia oleifera the leaf chlorophyll, soluble sugar and starch of terahertz spectroscopy response characteristics, extracting early recessive diseases of Camellia oleifera of leaf water conveyance terahertz imaging characteristic parameters, analysis of recessive diseases leaf terahertz image response mechanism, combined with chemometric methods, construct the recessive diseases of Camellia oleifera early rapid prediction stability mathematic model, and the model to verify the accuracy. The project intends to solve the long test cycle, high cost, professional and strong, it is difficult to ensure the inspection results and real-time condition consistency problems of traditional laboratory analysis. Provide technical support for the realization of project research recessive diseases of Camellia oleifera early prediction.

油茶产区当前普受病害影响严重，病害给油茶产业带来不可估量的经济损失，植物病害检测已成为国内外研究的热点。寻找快速、实时、准确和低成本的油茶病害早期诊断，实现及时预报和科学防治，保障油茶生产安全和促进农民增收具有重要意义。项目以油茶隐性病害为研究对象，利用LIBS光谱结合THz成像技术，开展油茶隐形病害早期快速诊断机理和方法研究。在寻找油茶隐形病害早期叶片营养元素的LIBS光谱变化规律的同时，研究隐形病害早期叶片叶绿素、可溶性糖和淀粉等生理指标的THz光谱响应特性，提取隐形病害早期叶片输水能力THz成像特征参数，分析隐形病害叶片THz图像响应特性，结合化学计量学方法，构建油茶隐性病害早期快速预测预报数学模型，并进行模型精度验证。项目拟解决传统检测周期长、成本高、专业性强、病理特征不清，难以保证检测结果与实时病情一致性等难点，为实现油茶隐性病害早期预测和科学防治提供理论与技术支撑。

油茶炭疽病是油茶最重要的病害之一，严重地影响了油茶产业的发展。因此及时、准确地检测炭疽病，对油茶产业健康持续发展具有重大意义。田间诊断、指示植物和实验室化学分析等油茶病害传统检测方法，具有耗时、费力、破坏性和检测指标单一等缺点，且主要是针对病害症状发生后的检测方法，已经无法满足油茶产业的实际需求。因此，亟需探索一种快速判别油茶早期病害的新方法，为实现油茶病害防治提供理论基础。. 项目开展了油茶叶片营养元素与油茶炭疽病相关关系研究，分析了样品中Fe、Mn、Cu、Zn、Ca和Mg六种营养元素的特征峰强度变化规律与炭疽病危害程度之间的关系；根据油茶叶片中Fe、Mn、Mg、Ca、Cu和Zn营养元素的LIBS光谱特征谱线的变化规律，从而确定早期油茶炭疽病的典型营养元素表征指标，建立基于全谱和基于特征变量的线性判别分析、极限学习机、支持向量机油茶叶片炭疽病判别分析模型。在基于LIBS技术定性判别分析的基础上，开展了基于LIBS结合THz技术的油茶叶片炭疽病等级研究，THz-LIBS光谱联合之后建模集精度和稳定性更高。采用塔姆黑尼（Tamhane）检验方法进行多重比较检验，进一步探究了各营养元素含量指标对不同炭疽病染病程度的影响。最后采用外标法、PLS和BP-ANN建立营养元素含量的定量分析模型。实现了对炭疽病油茶叶片中营养元素的精准测量和分析，提高了炭疽病早期快速诊断的准确率及效率。研究为油茶病害的早期快速判别和及时科学防治提供理论依据。 . 已发表论文10篇，其中8篇被SCI期刊检索，1篇被EI期刊检索，授权实用新型专利2项。