显色图像分析技术在水稻叶耳花青甙显色目测分级中的应用

Application of coloration image analysis technology in visual grading of anthocyanin coloration of rice leaf auricles

中文关键词:  水稻特异性、一致性、稳定性测试  叶耳花青甙显色强度  颜色特征  图像分析

英文关键词:rice distinctness, uniformity, stability test  anthocyanin coloration of leaf auricles  color features  image analysis

中文摘要:

      目的 基于颜色特征的叶耳花青甙显色分级研究。方法 以水稻倒二叶叶耳花青甙显色测试为切入点, 在Emgu Cv3.0图像分析软件基础上, 对完成图像分割的目标区域提取红绿蓝(red green blue, RGB)、色调饱和度亮度(hue saturation value, HSV)颜色特征, 利用SPSS软件对颜色特征和测试分级数据进行相关性、回归等统计分析, 建立颜色特征多元回归模型。结果 叶耳花青甙显色强度与R(红色)、G(绿色)、B(蓝色)、H(色调)、V(亮度)极显著负相关; 所有颜色特征值中, G值与叶耳花青甙显色强度相关性最显著, 是一元和多元回归主要自变量; G值建立的一元回归模型中, R2为0.980; 多元回归模型R2值为0.994。结论 回归模型的拟合效果好, 用这两个模型均可完成叶耳花青甙显色强度分级。

英文摘要:

      Objective To study the color grading of anthocyanin based on color characteristics. Methods Based on Emgu Cv3.0 image analysis software, the color test of anthocyanin in rice was taken as the starting point, the color features of red green blue (RGB) and hue saturation value (HSV) were extracted from the target area of the image segmentation, and the correlation and regression of the color features and test grading data were statistically analyzed with SPSS software to establish a multiple regression model of color features. Results Anthocyanin colorations of leaf auricle were significantly negatively correlated with R (red), G (green), B (blue), H(hue), V(value); among all the color features, the correlation between G value and anthocyanin colorations was the most significant, and moreover, G value was the major independent variable in univariate and multiple-variate regression; in G value-based univariate regression model, R2 was 0.980; in multiple-variate regression model, R2 was 0.994. Conclusion The fitting effect of regression model is good, and the color intensity classification of anthocyanin can be achieved by using these two models.