基于机器视觉的鸡蛋品质无损检测方法

Non-destructive testing method of egg quality based on machine vision

中文关键词:  鸡蛋新鲜度  机器视觉  图像处理  无损检测

英文关键词:egg fresh degree  machine vision  image processing  nondestructive testing

中文摘要:

      目的 以蛋品质量评价指标—哈夫单位值为检测标准, 建立一种基于机器视觉的鸡蛋品质无损检测方法。方法 通过摄像头捕捉到鸡蛋图像信息, 利用MATLAB对鸡蛋图像的G分量以及I分量进行特征参数提取, 并计算出与鸡蛋新鲜度相关的4个特征参数: 蛋黄面积比、气室面积比、气室高度比与气室直径比, 将其作为自变量, 通过高精度游标卡尺实测每个鸡蛋样本哈夫值作为因变量, 分别建立一元回归模型, 寻找特征参数与哈夫值的关系, 并根据哈夫值对鸡蛋新鲜度进行分级。结果 实验表明, 所测4个特征参数中, 蛋黄面积比与哈夫值存在较强的相关性, 相关系数为0.78, 拟合优度为0.62, 蛋黄面积比越小, 鸡蛋哈夫值越大, 说明鸡蛋越新鲜。结论 基于机器视觉的鸡蛋品质无损检测方法不仅具有较强的应用价值, 还可以为鸡蛋品质智能分级提供技术支撑。

英文摘要:

      Objective To establish a nondestructive testing method for egg quality based on machine vision wased, with the evaluation index of egg quality-Haff unit value as test standard. Methods Images of eggs were captured by a camera, than characteristic parameters in I component images and G component images of eggs were extracted based on MATLAB. Four characteristic parameters of egg freshness could be calculated, including tatio of egg yolk area, air chamber area, chamber height and chamber diameter. The single-element regression model was established with Haff values measured by high precision vernier caliper as dependent variables and the 4 characteristic parameters as independent variables to find the relationship between them respectively, and then the egg freshness was graded according to Haff values. Results The experiment showed that there was a strong correlation between the area ratio of egg yolk and Haff value. The correlation coefficient was 0.78 and the goodness of fit was 0.62. The smaller the egg yolk area ratio, the larger the egg Huff value, indicating that the egg was fresher. Conclusion The non-destructive testing method for egg quality not only has a strong application value, but also provides technical support for egg quality intelligent grading.