关键词: 无人机, 遥感, 果树, 氮素含量, 机器学习

Abstract: Nitrogen is an indispensable component for the growth and development of fruit trees. Nitrogen content beyond the normal range will affect the growth and development of trees, and directly or indirectly reduce the yield and quality of fruits. Therefore, rapid and accurate grasp of the nitrogen content of fruit trees can provide technical support for precise fertilization, so as to achieve high quality and high yield of fruit trees. With the rapid development of UAV industry, UAV remote sensing monitoring plays an important role in nitrogen content monitoring because of its advantages of non-destructive, fast, real-time and high efficiency. On the basis of introducing the current mainstream UAVs, this paper combs the data acquisition and subsequent processing methods, and expounds the monitoring of nitrogen content in fruit trees by multispectral, hyperspectral, visible light and other types of sensors. In general, multispectral and hyperspectral sensors have better effect on nitrogen monitoring of fruit trees, and the model constructed by machine learning method has higher accuracy than traditional methods. Finally, the deficiencies and future development direction of UAV remote sensing monitoring of fruit tree nitrogen content in four aspects of UAV flight platform and sensor performance, data acquisition and processing, promotion and application, and use policy were put forward, so as to provide reference for the accuracy, efficiency and intelligence of UAV remote sensing monitoring of fruit tree nitrogen content in China.

Key words: UAV, remote sensing, fruit tree, nitrogen content, machine learning