Diagnosis method for corn nitrogen status measurement based on multispectral imaging

Monitoring the nitrogen status of corn crop has significant influence in rational fertilization, production expansion, and pollution (caused by excessive use of nitrogen fertilizer) reduction. In order to rapidly acquire corn nutrient information in the field, a nondestructive method of corn nitrogen content measurement was conducted based on multispectral imaging. The multi-spectral image monitoring system was deployed to acquire the canopy images of corn at jointing stage, and each sample's nitrogen content was measured by AA3 continuous flow analyzer (CFA). 11 vegetation indices, including AVS R , AVS G , AVS NIR , NDVI, GNDVI, RVI R/G , RVI R/NIR , RVI NIR/R , RNR, GNR and NIRNR, were calculated. Correlation analysis was then conducted to reduce the dimension of data so as to acquire 3 sensitive spectral characteristic parameters. The corn nitrogen nutrition diagnosis model at jointing stage was built with simple linear and polynomial regression method based on single sensitive spectral characteristic parameter. Moreover, the model was built with least squares multiple regression method with 3 sensitive spectral characteristic parameters. It was feasible to diagnose nitrogen content of corn based on multi-spectral images.