Estimation of chlorophyll content in radish leaves using hyperspectral remote sensing data and machine learning algorithms

Slag consists mostly of mixed oxides of elements such as silicon, and recycled slag may be used for cultivating. The relationship between slag fertilization and plant growth rate can be expected to change depending on the volume of slag applied. Quantifying the chlorophyll content, an effective indicator of disease as well as nutritional and environmental stresses on plants, will enable optimal slag fertilization and then monitoring chlorophyll content using field measurements would enable the determination of optimal slag fertilization rates. In this study, radish plants (Raphanus sativus L), which belongs to the family Brassicaceae and is popular root vegetable in both tropical and temperate regions, were cultivated with slag fertilization and the potential use of hyperspectral reflectance was evaluated. Some preprocessing techniques were effective for retrieving chlorophyll contents in radish leaves from hyperspectral reflectance and then the regression model based on random forest and continuum-removed reflectance had the highest performance with a root mean square error of 5.141 mu g cm(-2) and RPD values of 1.858 for the test data set.