Abstract: By measuring the hyperspectral reflectance of maize plant leaves with different types of microorganism under the condition of pot experiment, the difference of the leaf water content was analyzed,Investigate the effect of different microorganisms on the water content of vegetation and the change of its spectral characteristics,and tried to use hyperspectral remote sensing technology to monitor and identify the water content of maize leaves under different treatment conditions. The results show that the CK prediction model Y=0.839×[exp（0.006 t）] has the highest prediction accuracy in the best estimation model of the leaf water content, the coefficient of determination is 0.851,RMSE is 0.018,the prediction precision of G.m processing model Y=0.84-100.128 t-166 349.654 t² is the lowest, the coefficient of determination R² is 0.638,RMSE is 0.013. The models constructed by different treatments can effectively predict the water content of the corresponding leaves,in the field of microbial reclamation,information support is provided for the rapid and efficient monitoring and evaluation of the water content of vegetation leaves under the action of different types of microorganism.

Key words: hyperspectral remote sensing, microbial reclamation, leaf water content, estimation model