Nutritional Diagnosis of Camellia oleifera Using UAV Hyperspectral Imaging and ES-CNN Strategy
Camellia oleifera,an important woody oil crop in China,is significantly influenced by its nutritional status in terms of yield and quality.Traditional nutritional diagnosis methods for Camellia oleifera are time-con-suming,labor-intensive,and challenging for large-scale monitoring.This study aims to develop a rapid,non-de-structive,and accurate nutritional diagnosis method for Camellia oleifera using Unmanned Aerial Vehicle(UAV)hyperspectral remote sensing technology,providing technical support for precise fertilization.A typical Camellia oleifera planting area was selected as the research site.A drone equipped with a hyperspectral imager was used to acquire Hyperspectral Images(HSI)of the Camellia oleifera canopy.The canopy area was extracted through ob-ject-oriented classification,and spectral analysis was conducted based on GPS point sampling(GPS Sample,GS)and Ecognition software classification results(Ecognition Sample,ES).After dimensionality reduction using Prin-cipal Component Analysis(PCA),a Convolutional Neural Network(CNN)algorithm was employed in conjunction with ground measurements of key nutritional elements(nitrogen,phosphorus and potassium)in Camellia oleifera leaves to evaluate content accuracy.The results show that the spectral characteristics of the Camellia oleifera can-opy enable accurate extraction.The estimation model of nutrient element content in Camellia oleifera leaves based on PCA components has high accuracy and can effectively diagnose the nutritional status of Camellia oleifera.The strategy combining ES sampling and CNN(ES-CNN)achieved higher accuracy due to more representative sam-pling data.The detection models for nitrogen,phosphorus,and potassium showed high accuracy,with R2 values of 0.63,0.52,and 0.53 respectively,and Root Mean Square Error(RMSE)values of 0.11,13.07,and 7.78 respectively.In conclusion,UAV hyperspectral imaging technology combined with the ES-CNN strategy provides a fast,non-destructive,and accurate new method for nutritional diagnosis of Camellia oleifera.This approach can guide pre-cise fertilization,enhance yield and quality,and promote the sustainable development of the Camellia oleifera in-dustry.
NETL
NSTL
万方数据
