小麦长势无人机遥感监测关键指标参数反演机理及长势诊断模型研究

TThe real-time monitoring of wheat growth can provide information support for wheat fine management. The UAV has the advantage of real-time and dynamical data acquisition and high resolution, and it will be the trend to apply the UAV remote sensing to monitor the wheat growth in future. At present, the real-time monitoring of wheat growth depends mainly on a single parameter, which can not reflect the size and physiological activity of population at the same time. It is also difficult to fully reflect the actual growth status. Based on UAV remote sensing information, the leaf area index, biomass, chlorophyll content and nitrogen content are selected as the key parameters, and then they are normalized as a comprehensive growth index, which can be used as the key index for evaluating the wheat growth. Research contents: 1. Through sensitivity analysis, the sensitive bands are selected to construct the spectral index; combined with the wheat growth mechanism, the different spectral characteristics of growth indicators are explored to study the inversion mechanism of the key wheat growing indicators; analyzing the correlation between growth indicators and spectral index, the index inversion models are constructed and then the accuracy of the models are verified; 2. Comparing the effect of different growth indexes on wheat population structure and physiological activity, the appropriate indicators are selected to constructed the wheat growth status diagnosis models; the classification standards of different growth conditions are set up to classify the wheat growth status and achieve the quantitative evaluation of wheat growth. The research results will perfect the system of wheat growth monitoring and improve the accuracy of growth evaluation.Further more, the practical application value will be expanded.

小麦长势监测，可以为小麦田间精细管理提供信息。发挥无人机遥感机动灵活、实时获取数据和高分辨率优势，进行小麦长势监测，将成为未来发展趋势。目前小麦长势监测主要依靠单一指标，不能同时反映群体结构大小和生理活性高低，难以全面反映实际生长状况。项目基于无人机高光谱遥感信息，选择叶面积指数、生物量、叶绿素和氮含量等指标，并归一化为综合长势指标，作为评价小麦长势关键指标。拟研究内容：① 通过敏感性分析，筛选敏感波段，构建光谱指数；探究不同长势指标波谱特性，结合小麦生长机理，研究小麦长势关键指标反演机理；分析长势指标与光谱指数的相关性，构建指标反演模型，对模型精度进行验证；② 比较分析不同长势指标对小麦群体结构与生理活性的影响，筛选适宜指标，构建小麦长势状况诊断模型；制定不同长势状况分级标准，划分小麦长势状况等级，实现小麦长势定量评价。研究成果将完善小麦长势监测方法，提高长势评价精度，扩展实际应用价值。

对农作物进行长势实时监测和长势定量评价，可以促进农业生产科学管理，为各级政府部门和广大农民提供重要的科学决策依据。本研究以冬小麦为研究对象，在无人机高光谱数据获取与预处理的基础上，重点研究了如下内容：（1）提出多目标无人机微型凝视高光谱成像仪辐射校正算法，通过比较定标后的高光谱图像光谱与地面光谱仪实际测量的地物光谱来验证该方法的辐射定标的精度；（2）研究了小麦不同生育期和不同氮胁迫处理水平下冠层光谱特征，为小麦长势监测关键参数遥感反演奠定基础；（3）提出基于分数阶微分和连续小波变换的小麦叶面积指数、生物量、叶绿素含量和氮含量遥感反演方法，在无人机高光谱数据分数阶微分处理和连续小波变换的基础上，构建分数阶微分光谱和小波变换系数，进行相关性分析，基于机器学习算法，构建不同生育期叶面积指数、生物量、叶绿素含量和氮含量遥感反演模型；（4）提出基于综合指标参数的小麦长势监测方法，将叶面积指数、生物量、叶绿素含量和氮含量归一化处理，构建长势综合指标参数，实现小麦长势实时诊断和定量评价。为制定小麦田间实时管理措施与指导精准作业提供信息支撑和科学依据。