Simultaneous and nondestructive diagnostics of nitrogen/magnesium/potassium-deficient cucumber leaf based on chlorophyll density distribution features

Nitrogen (N), magnesium (Mg), and potassium (K) deficiency all result in the decrease of chlorophyll density (CH.D) in cucumber leaf, but it is difficult to differentiate N/Mg/K-deficient leaves using CH.D. New nutrient indicators based on CH.D distribution features of the whole cucumber leaf were proposed to simultaneously diagnose N/Mg/K-deficient cucumber plants. CH.D distribution map in N/Mg/K-deficient and control leaves were determined non-destructively using hyperspectral imaging with the wavelength range of 430–960 nm. Compared with the CH.D distribution map of control group leaves, a decrease in CH.D could be observed on the whole area of N-deficient leaves, Mg-deficient leaves presented pixels with low CH.D between main leaf vein, and pixels with low CH.D at the border of K-deficient leaves. Distribution features were extracted from the CH.D distribution map and were employed to diagnose N/Mg/K-deficient leaves simultaneously, with the diagnostics rates of 96.67%. The results indicated that CH.D distribution features can realize the simultaneous diagnosing in the early stage of N/Mg/K deficiency and timely guidance of top dressing is of great significance to increase yield.