论文部分内容阅读
作物病害表现在叶片形态上,且外观和内部结构发生变化,生长环境也对病害有一定的影响.将生长环境、叶RGB图像和光谱图像融合,研究并提出一种作物光谱图像相关模型的时空信息挖掘方法,从时间维度、空间维度和光谱维度分析作物病害的光谱反射特征与作物发育、健康状况和生长条件的相关性,建立典型病害特征模型.实验结果表明,图像处理和光谱成像技术的融合方法可以在疾病的早期阶段实现快速、准确和无损诊断.“,”When crop disease occurs,it is often displayed in the leaf,and the appearance and internal structure of the crop are changed,and the growth environment also has a certain influence on the disease.The growth environment,leaf RGB images and spectral images are fused to study the sparse feature recognition method of crop diseases based on information combination of multi spectral images.In this paper,a spatial-temporal information mining method for crop spectral and image correlation models is studied.The correlation between spectral reflectance characteristics of crop diseases and crop development,health status and growth conditions are analyzed from time dimension,space dimension and spectral dimension,and disease characteristics is established.The experimental results show that the fusion method of image processing and spectral imaging technology can achieve fast,accurate and nondestructive diagnosis in the early stage of disease.