Estimation of leaf chlorophyll content of rice using image color analysis

摘要

Leaf color has been commonly used as an index for crop stress status diagnosis. We have developed a low-costnand nondestructive method that is easy to use to evaluate the chlorophyll content of rice (Oryza sativa L.) using leaf imagencolor analysis. The relationships between the imaging data and leaf pigment content were investigated. There was a highlynsignificant negative relationship between the red (R) and green (G) values in the RGB color space analyzed from leafnimage and chlorophyll a (Chl a) content, chlorophyll b (Chl b) content, Chl au0002b content, and carotenoid (CAR) content.nThe G data had a higher correlation coefficient with Chl a, Chl b, Chl a u0002 b, and CAR than the R data. However, nonsignificant relationship was found between the blue (B) value and Chl a, Chl b, Chl a u0002 b, and CAR. Linear andnlogarithmic correlation functions were used to model the relationship between imaging data and leaf pigment data. Usingnanother set of collected data, significant correlations were observed between the predicted Chl a, Chl b, Chl a u0002b, CARbasednG data and the measured values. The determination coefficient of the R predicted model of simulated chlorophyllnpigment content and observed data was smaller than that of the G predicted model. Comparably, Chl a u0002 b and Chl ancould be better predicted than Chl b and CAR from rice leaf image analysis. Combined with our previous study on barleynand wheat, this study demonstrated that the chlorophyll content of plants could be nondestructivly evaluated using thenleaf image color analysis method.