基于智能手机水稻叶色的最佳测试位置的确定及RGB值与SPAD值的关系
Determination of the Optimal Testing Location for Rice Leaf Color Based on Smartphones and the Relationship between RGB Value and SPAD Value

为探明水稻叶色的最佳测试位置，最佳的颜色特征参数，并构建颜色特征参数与SPAD值的关系模型。利用叶绿素计(SPAD-502型)获取叶片的SPAD值及智能手机获取的颜色空间RGB值，分析叶片各测试位置的SPAD值之间的关系及颜色特征参数与SPAD值之间的关系。结果表明，水稻叶片中间1/2及上下3 cm处的SPAD值的平均值与整片叶的SPAD值平均值相关系数最高；水稻叶片两侧SPAD值的平均值、标准差、变异系数的大小分别为，光滑侧 > 粗糙侧、光滑侧 < 粗糙侧、光滑侧 < 粗糙侧，正背面接近1:1关系；颜色特征参数G-B值与SPAD值的相关性最好，并构建函数关系模型，Y_SPAD = ？0.4541X_G-B + 64.4618 (n = 48, df = 45, R² = 0.7475, F = 133.2052, p < 0.01)。综上，基于智能手机的颜色识别器测量水稻叶色，能为代替价格昂贵、性价比低、操作繁琐等仪器的水稻氮营养监测提供新方法。
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, R² = 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.