基于多角度成像高光谱的生姜氮素垂直分布反演机理与建模研究

Ginger (Zingiber officinale) is an important economic crop with medicine and food.function, which has high developing and utilizing values. Nitrogen is very.important for the growth of ginger. However, excess fertilization will lead to the.growth of stem and leaves, which effects the fertilizers distribution of tubers.and brings a decrease in the yield. As remote sensing is a quick and.non-destructive way for obtaining the object information, it has been widely used.in crop N status diagnosis. But these data are often spectral reflectance or.digital images by observing vertically, which ignored the N vertical distribution,.more importantly, the N accumulation amount of crops have been seldom studied..Therefore, this study will focus on the modeling and quantitative inversion.mechanism of ginger nitrogen vertical distribution by using multi-angle.hyperspectral imaging data. The objectives include: (1) building the mathematical.models of ginger N accumulation and vertical distribution; (2) studying the.inversion mechanism of ginger N vertical distribution by multi-angle hyperspectral.imaging technique; (3) building the monitoring models for ginger N vertical.distribution which combined the agronomy with remote sensing. Based on the.above-mentioned analysis, the mechanism between the ginger N vertical distribution.and N accumulation amount in the roots, stems and leaves will be explored,.meanwhile, the application of multi-angle hyperspectral imaging technique in.different crops will also be expanded, this will be of important significance to.modern agricultural production and decision making.

生姜作为一种药食两用的经济作物，具有很高的利用开发价值。氮素对生姜的生长发育至关重要，但过多施用氮肥，会促进地上茎、叶生长，影响氮素向块茎的分配，导致减产。遥感技术具有快速、无损的获取地物信息的特点，在作物氮营养诊断中得到了广泛的应用，但现有的遥感数据多为单一的地面高光谱和数字图像，观测的方式也多为垂直观测，忽略了作物氮素垂直分布的特性，更缺乏对作物氮累积进程的探讨。本项目采用多角度成像观测技术，开展生姜氮素垂直分布的定量反演与建模的研究，内容包括：（1）建立生姜氮素累积进程及垂直分布的数学模型；（2）研究生姜氮素垂直分布的多角度成像高光谱遥感反演机理；（3）建立融合农学知识与遥感机理的生姜氮素垂直分布遥感监测模型。通过本项目的研究，阐明了生姜根-茎-叶氮素累积与垂直分布的规律，拓展了多角度成像光谱技术在不同作物中的应用，对于现代农业生产和决策具有十分重要的科学意义。

生姜作为我国现代农业特色蔬菜产业体系的重要作物，其氮营养的快速无损诊断对现代农业生产和决策具有十分重要的科学意义。现有的作物氮素遥感反演多采用高光谱垂直观测方式，忽略了作物氮素垂直分布特征。本项目通过2年田间小区试验，开展的研究和取得的成果如下：（1）生姜氮素垂直分布规律为：幼苗期和旺盛生长期上层叶片氮含量高于中、下层，根茎膨大期和收获期，下层氮含量最高可达2.90%；（2）利用成像光谱仪提取各个叶层叶片的光谱反射率，植被指数MCARI[705, 750]/OSAVI [705, 750]对上、中、下层叶片氮含量反演效果最好。经波段组合后，NDSI(752, 751)对生姜下层叶片反演精度R2最高达0.91；（3）利用连续小波变换的方法对生姜各个叶层叶片氮含量进行反演，当分解尺度为3，波段为561nm时，各叶层氮含量反演精度R2分别为85.10%、91.42%和93.40%；（4）利用实测的多角度高光谱数据和ACRM辐射传输模型，构建了生姜冠层氮含量（CNC）垂直分布的指数反演模型，得出后向观测反演效果优于前向观测。通过对模拟和实测数据的分析，当观测角度为-20°时，DDn对第1、2层CNC反演精度最高。经波段组合后，SR(484, 472)对第3层CNC反演精度R2最高可达0.92。上述研究结果，有助于将作物氮含量无损诊断技术引入到经济作物中，所采用的生姜氮素垂直分布多角度遥感反演方法，有助于提高反演精度，为生姜的精准施肥提供技术支持。