基于高光谱和电子鼻检测水果采后真菌病害的信息基础和机理研究

Fungal diseases are one of the main reasons causing fruit decay which may lead to damage of fruit quality and safety during postharvest, and it is very important and necessary to detect and monitor fruit fungal diseases nondestructively. Peach and strawberry are two fruits which are easy to become decay during post-harvest storage and transportation, and the aims of this research are to detect fruit quality and safety based on changes of fruit texture, microorganisms, flavor parameters by measurement of hyperspectral imaging and electronic nose. From the information measured by hyperspectral imaging and electronic nose, the relationship between hyperspectral imaging and electronic nose and those nondestructive determined parameters can be established. Different fungal diseases and quality decay can be nondestructively detected by analyzing the results and possible detecting mechanism will be explored. Further data analysis of the two nondestructive measurements can combine the information of image and flavor parameters, and can enable us to establish detection models for fungal diseases and possible fungal identification by hyperspectral imaging and electronic nose. Information and mechanism of fruit decay caused by fungal diseases can be form the base for predicating fruit quality and safety based on above mentioned two nondestructive testing methods. The results from this research can be applied to develop equipment for testing and monitoring fruit quality and safety during postharvest storage and transportation of peach and strawberry fruits. It can also provide useful information for fruit fungal diseases detection and monitoring of decay sensitive fruits in postharvest storage and transportation.

真菌侵染是造成水果采后腐败的重要原因，对水果采后质量安全造成严重危害。因此，对水果采后真菌病害的快速无损检测和监测显得十分重要。项目以采后易腐水果桃和草莓为主要研究对象，通过分析水果采后真菌病害造成的果实品质、微生物指标、挥发物以及高光谱图像特性、电子鼻响应等信息的变化和差异，在确定果实真菌病害和腐败程度的表征性参数和范围后，探明反映果实真菌病害的表征性参数信息，对应的高光谱图像和电子鼻响应，确定高光谱图像和电子鼻检测病害的特征参数，并进一步对高光谱图像和电子鼻信息进行数据层和特征层融合，探索和构建高光谱图像和电子鼻检测水果采后真菌病害和致病真菌类型的较佳判别模型，明确高光谱图像和电子鼻对水果采后真菌病害检测的机理，建立易腐水果贮运中真菌病害的监测、预测模型。研究可为水果采后病害自动化检测技术和设备开发提供基础数据和理论参考，也为易腐水果采后贮藏和流通提供基于信息融合无损检测的基础。

真菌侵染是造成水果采后腐败的重要原因，对水果采后质量安全造成严重危害。因此，对水果采后真菌病害的快速无损检测和监测显得十分重要和关键。项目以采后易腐水果桃和草莓为主要研究对象，通过分析水果采后真菌病害造成的果实品质、微生物指标、挥发物以及高光谱图像特性、电子鼻响应等信息的变化和差异，在确定果实真菌病害和腐败程度的表征性参数和范围后，探明反映果实真菌病害的表征性参数信息对高光谱图像和电子鼻响应的影响和相互联系，确定高光谱图像和电子鼻特征参数，并进一步对高光谱图像和电子鼻信息进行数据层和特征层融合，探索和构建高光谱图像和电子鼻检测水果采后真菌病害和致病真菌类型的较佳判别模型，明确高光谱图像和电子鼻对水果采后真菌病害检测的机理，建立相关的检测、预测模型。结果表明：桃和草莓果实采后受到不同典型真菌侵染，表现出不同的高光谱和气味特征，依据这些特征可以进行品质和病害判定。利用光谱特征建立的模型对桃病害与健康果实的整体判别率达到95.5%，对不同病害类型的识别达到84.1%；基于电子鼻信号的混合病害类型识别率达到85.7%；对早期病害构建的数据融合模型，达到最佳预测准确率为97.5%。基于特征光谱建立的草莓病害程度的SVM模型，获得94.4 %的准确率；根据电子鼻信号构建草莓早期病害分类模型预测准确率为92.9 %；而根据多源特征数据融合的草莓采后不同真菌早期病害鉴定的识别达到96.8 %，优于单一技术。真菌感染组在贮藏过程中水分、糖分 叶绿素等的变化与无侵染组显著不同，对应着特征光谱的变化；特征气味有明显区别，对应的气味传感器有不同的响应。通过研究明确了如何确定这些特征参数，并依据优化的参数与建模方法建立判别与预测模型，提出了基于数据融合的判别和预测桃、草莓早期病害判别的方法和模型。本研究可以为水果采后病害无损检测技术和设备的开发提供基础数据和参考，也为水果采后贮藏和流通领域的质量安全提供了理论依据。