Hyperspectral imaging techniques for rapid detection of nutrient content of hydroponically grown lettuce cultivars

The analyses of online quality measurements four lettuce cultivars (Rex, Tacitus, Black Seeded Simpson, Flandria) using hyperspectral image processing techniques have been studied. Seedlings were planted in Rockwool cubes and fed for 3 weeks hydroponic nutrient solution containing 0, 50, 100, 150, 200, 250, 300 ppm nitrogen. images the freshly cut leaves captured camera wavelength range 400–1,000 nm. algorithm to measure levels leaf tissues including nitrate (NO3–), calcium (Ca2+), potassium (K+), soluble solid content (SSC), pH, total chlorophyll concentration (SPAD reading) was developed. To locate optimum wave band combination values, simple linear regression models with two spectral data namely reflectance its first derivative, indices, ratio index (RSI) normalized difference (NDSI), found. optimal bands within 506–601 nm 634–701 NO3–, Ca2+, K+, SSC, SPAD samples estimated partial least squares (PLSR) principal component analysis (PCA) bands. Both PLSR PCA produced well correlated outcome comparison laboratory measured predicting contents R2 = 0.784–0.987.