基于计算机视觉技术的番茄花青素含量检测.docxVIP

农机化研究 2018,03(40), 193-197 基于计算机视觉技术的番茄花青素含量 检测 雷静 成都农业科技职业学院 导出/参考文献关注 分享 收藏打印 摘要： 花青素是一种天然色素，营养保健价值极高。番茄可以同时作为蔬菜和水果，但 其花青素的含量很低，因此富含花青素的功能型番茄便成为一个重要的育种R 标。花青素的含量测定通常采用常规的分光光度法或色谱法，都有一定的局限 性。目前，计算机视觉测定各种色素含量的技术已经成熟，但应用于花青素的报 道还很少。为此，利用计算机视觉技术对番茄图像进行分析，从18个颜色特征 参数中发现了 6个与花青素实际含量之间存在极显著相关性。分别以这6个参数 为函数，建立了两者Z间的一元二次方程的回归模型，通过决定系数和均方根 误差评价，发现以r和R/ (G+B)为特征参数建立的模型拟合度最好，对检验集 的预测表现出很高的精确度，可以用于花青素含量的快捷无损测定。 关键词： 汁算机视觉;番茄;花青素;含量测定; 作者简介：雷静(1981-),女，成都人,讲师，硕士，(E-mail) lpyppg@163. com。 收稿日期：2017-01-11 基金：四川省教育厅项目仃3ZA0335) Study on the Detection of Tomato Anthocyanin Content Based on Computer Vision Technology Lei Jing Chengdu Agricultural College; Abstract： Anthocyanin is a natural pigment, nutritional health value is very high. Tomato can be used as a vegetable and fruit at the same time, but its content of anthocyan in is very low. Therefore, functional tomato rich in emthocyanin has become an importsnt breeding target. Determination of anthocyanin contcnt usually by conventional spectrophotometry or chromatography, have some limitations. At present, computer vision measurement of various pigment content of the technology has matured, but the application of anthocyanins in the report is still very small. In this paper, the tomato image was analyzed by using computer vision technology- The resuIts showed that there were significeint correlations between the anthocyanin content and the content of 6 anthocyanins from 18 color characteristic parameters. Using these six parameters as a function, a regression model of the quadratic equation between the two is established. Through the evaluation of coefficient and root mean square error, it is found that the model fitting based on r and R/ ( G + B) is the best，and the prediction of test set shows high accuracy- Fast and Nondestructive Determination of the Content of. Keyword： computer vision; tomato; anthocyanin; determineition of content; Received： 2017-01-11 0引言 花青素是一种类黄酮类的水溶性色素，广泛存在于植物体中。植物细胞液为酸性 吋花青素呈现红色，为碱性吋则呈现蓝色。花青素是植物的一种次生代