Determination of the Optimal Testing Location for Rice Leaf Color Based on Smartphones and the Relationship between RGB Value and SPAD Value

Abstract

To identify the optimal testing location and color feature parameters for rice leaf color, and to construct a relationship model between color feature parameters and SPAD values, a chlorophyll meter (SPAD-502 type) was used to obtain the SPAD value of leaves and the RGB value of color space obtained by a smartphone to analyze the relationship between SPAD values at various test positions of leaves and the relationship between color feature parameters and SPAD values. The results showed that the correlation coefficient between the average SPAD values at the middle 1/2 and upper and lower 3cm of rice leaves and the average SPAD values of the entire leaf was the highest. The average value, standard deviation, and coefficient of variation of SPAD values on both sides of rice leaves are as follows: smooth side > rough side, smooth side < rough side, smooth side < rough side, with a close 1:1 relationship between the front and back sides. The correlation between the color feature parameter G-B value and SPAD value is the best, and a functional relationship model is constructed, Y SPAD = − 0.4541X G-B + 64.4618 (n = 48, df = 45, R 2 = 0.7475, F = 133.2052, p < 0.01). In summary, color recognition devices based on smartphones can provide a new method for monitoring rice nitrogen nutrition, replacing expensive, cost-effective, and cumbersome instruments.