中文摘要：

为了实现可靠的植物病虫害诊断,文章提出把光谱分析技术和多光谱成像技术相结合的方法用于常见的红粉、黑星、白粉、褐斑和霜霉五种黄瓜病害的识别研究。实验采用窄带多光谱成像技术在标准观测环境下获取患病黄瓜叶面的14个可见光通道和近红外通道、全色通道的多光谱图像。利用距离法、光谱角度匹配法和相关系数法对病斑样本的光谱信息进行学习分类。对于实验提出的七种分类情况,通过距离法和相关系数法组合筛选,对红粉、白粉、黑星和白板的分类正确率达100%,对霜霉及无病的分类正确率达80.00%和93.33%,平均分类正确率为81.90%。实验结果表明,光谱分析方法和多光谱成像技术结合能全面、快速、精确提取植物病害的信息,实现分类,为对植物病害进行快速、准确和非破坏性诊断提供技术支持。

英文摘要：

For a reliable diagnosis of plant diseases and insect pests, spectroscopy analysis technique and mutispectral imaging technique are proposed to diagnose five cucumber diseases, namely Trichothecium roseum, Sphaerotheca fuliginea, Cladosporiurn cucumerinum ,Corymespora cassiicola and Pseudoperonospora cubensis. In the experiment, the cucumbers＇ multispectral images of 14 visible lights channels, near infrared channel and panchromatic channel were captured using narrow-band multispectral imaging system under standard observation environment. And the 5 cucumber diseases, healthy leaves and reference white were classified using their multispectral information, the distance, angle and relativity. The discrimination of Trichotheeium roseum, Sphaerotheca fuliginea, Cladosporiurn cucumerinum, and reference white was 100%, and that of Pseudoperonospora cubensis and healthy leaves was 80% and 93.33% respectively. The mean correct discrimination of diseases was 81.90% when the distance and relativity were used together. The result shows that the method realized good accuracy in the cucumber diseases diagnosis.