为探索无人机遥感在玉米监测中应用效果和能力，本文以2018和2019年在河南省新乡县中国农业科学院农田灌溉研究所试验基地的玉米为研究对象，利用八旋翼无人机搭载的MicaSense RedEdge多光谱相机对试验区进行遥感监测，构建了玉米叶面积无人机遥感监测模型和产量估算模型，并在示范区进行了应用。结果显示，NDVI、EVI和 GNDVI三种植被指数在构建叶面积指数监测模型中具有较好的精度和稳定性。利用抽雄期植被指数构建的估产模型精度最高、吐丝期次之，拔节期最低，与单生育期估产模型相比，累积3个生育期植被指数构建的估产模型精度有一定提升，R2为0.87，RMSE为405.42 kg·hm-2。该研究构建的无人机遥感监测模型，可以快速有效评估玉米长势和产量，为玉米无人机遥感精准监测相关研究积累了经验，对规模化精准农业管理调控具有重要意义。
英文摘要:
The objectives of this study were to explore the application effect and ability about the Unmanned Aerial Vehicle (UAV) remote sensing in maize (Zea mays L.) growth monitoring. The experiments were carried out during 2018~2019 at Experimental Bases Farmland Irrigation Research Institute, Chinese Academy of Agricultural Sciences, Xinxiang, China, and study the estimate models of leaf area index and yield of maize derived from Unmanned Aerial Vehicle (UAV) multispectral remote sensing in different growth stages, and then mapping the leaf area index and yield of maize in demonstrative?region based on these models. The result?showed that: there were accuracy and stability of the estimate model about leaf area index based on NDVI, EVI and GNDVI. There was the highest accuracy of the yield estimate model based on vegetation indexes in tasseling stage, followed by the silking stage, and the lowest jointing stage. Compared the accuracy of the yield estimation model, the cumulative vegetation index from 3 growth stages was higher than the single growth stage, and the determination coefficient (R2) is 0.87, root mean square error (RMSE) is 405.42 kg·hm-2. The monitoring model of UAV remote sensing in this research, could monitoring and evaluation for growth potential and yield of maize by quickly and effectively, which has accumulated experience for precision agriculture based on UAV, and it is great significance for the management and control of precision agriculture on a large scale.
;
