Stripe rust is one of the epidemic diseases that cause massive yield reduction of wheat. The rapid monitoring of wheat stripe rust is of great significance for its control. Hyperspectral remote sensing has been increasingly used in the identification and prediction of crop diseases. In order to determine the appropriate method for monitoring wheat stripe rust by hyperspectral remote sensing, this paper calculates multiple hyperspectral features and solar-induced chlorophyll fluorescence (SIF) characteristics based on canopy measured hyperspectral and radiance. Two feature selection algorithms, variable importance in projection (VIP) and k-neighbor mutual information(k-MI), were used to screen the hyperspectral characteristics and SIF features which are sensitive to wheat stripe rust. The selected feature variables were used as input variables of partial least squares (PLS), extreme learning machine optimized by particle swarm optimization (PSO-ELM) kernel extreme learning machine optimized by particle swarm optimization (PSO-KELM) to construct a wheat stripe rust prediction model. The study shows that the number of features selected by k-MI is less than that of VIP method. Between the models constructed by the optimal features of the two methods, the prediction accuracy of the PSO-KELM model is the best, and the PSO-KELM model constructed by the k-MI method is the most accurate in which the training set R2 is 0.951, the verification set R2 is 0.952, and their corresponding root mean square errors (RMSE) are respectively 0.070 and 0.066. Conclusively, compared with the PLS model and the PSO-ELM model, the PSO-KELM model improves the stability and generalization ability of the model. It can be used as a new method for identifying the incidence of stripe rust and provides new ideas for further realization of large-scale high-precision remote sensing monitoring of crop health status. Keywords: wheat; stripe rust; Hyperspectral; solar-induced chlorophyll fluorescence; feature selection; kernel extreme learning machine; particle swarm optimization;
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.