Corn chlorophyll content detection based on spectral reflectance inversion absorptance

The absorption of light energy is an important indicator for evaluating chlorophyll content. Due to the complexity of absorptance collection and the simplicity of reflectance collection, currently commonly used spectral detection equipment in the field collects crop leaf or canopy reflectance instead of absorptance. However, the transmitted light energy of crops complicates the relationship between reflectance and absorptance, and the correlation between reflectance and chlorophyll content becomes weaker. Therefore, this study proposed a method to convert the reflectance of crop leaves into absorptance according to Beer-Lambert law for a more accurate prediction of chlorophyll content. First, the reflectance and transmittance of corn leaves are obtained simultaneously from hyperspectral images to calculate the absorptance. Then, the partial least squares regression models were established by reflectance, transmittance, and absorptance of full wavelength with chlorophyll content. In the reflectance model, R²C is 0.86, R²V is 0.70, root mean square error of calibration set (RMSEC) and RMSEV are 2.95 and 3.73, respectively. In the transmittance model, R²C is 0.86, R²V is 0.71, root mean square error of calibration set (RMSEC) and RMSEV are 2.93 and 3.66, respectively.The reflectance was converted into absorptance and cconstructed a new model. In the new model, R²C is 0.86, R²V is 0.78, RMSEC and RMSEV are 2.91 and 3.50, respectively. The results show that the R²V is increased by 0.08 and the error is reduced by 0.23 in the new absorptance model. Therefore, converting spectral reflectance into absorptance can more accurately assess chlorophyll content and provide technical guidance for precise crop management.