基于无人机多光谱图像的荔枝叶片氮磷钾含量检测

Detection of Nitrogen,Phosphorus and Potassium Content in Litchi Leaves Based on UAV Multispectral Image

利用搭载在无人机上的多光谱相机,以荔枝树冠层叶片作为试验对象,在光线充足且无阴影遮挡的环境下对荔枝树叶片样本进行拍摄,结合拍摄获得的多光谱图像数据和荔枝树叶片氮、磷、钾实际含量,通过相关性分析分别筛选与荔枝树叶片氮、磷、钾含量相关性较强的特征波段。根据相关特征波段的数据和荔枝树叶片实际含量,利用最小二乘法对所得数据进行回归分析,根据分析结果建立荔枝树叶片氮、磷、钾含量反演模型。结果表明,利用多光谱图像光谱信息并结合荔枝树叶片氮磷钾含量建立的反演模型,得到相关系数分别能达到0.8853、0.9116和0.9045。结果表明,通过此方法可以实现对荔枝树叶片氮、磷、钾元素含量的快速检测,帮助果农实时监测荔枝树的营养状况,并为精准施肥提供依据。

Using the multispectral camera mounted on the UAV,taking the leaves of the lychee tree canopy as the test object,the samples of litchi tree leaves were photographed in an environment with sufficient light and no shadows.The actual contents of nitrogen,phosphorus and potassium were screened by correlation analysis to screen the characteristic bands with strong correlation with the contents of nitrogen,phosphorus and potassium in litchi leaves.According to the data of the relevant characteristic bands and the actual content of litchi leaves,the least squares method was used to carry out regression analysis on the obtained data,and the inversion model of nitrogen,phosphorus and potassium content in litchi leaves was established according to the analysis results.The results show that the correlation coefficients can reach 0.8853,0.9116 and 0.9045 respectively by using the spectral information of multi-spectral images combined with the inversion model established by the content of nitrogen,phosphorus and potassium in litchi leaves.The rapid detection of phosphorus and potassium content helps fruit farmers monitor the nutritional status of lychee trees in real time and provides a basis for precise fertilization.