鸡蛋新鲜度综合无损检测模型及试验

为了提高鸡蛋新鲜度无损检测的准确性，该文用反映鸡蛋内部颜色的计算机图像参数和鸡蛋敲击响应特性参数构成鸡蛋新鲜度综合无损检测模型，利用两种检测方法之间的参数进行相互补偿和修正。用机器视觉装置获取鸡蛋颜色的亮度、蛋形指数，用声音采集装置获取声音的功率谱面积、共振峰频率、X轴方向的质心，通过试验测得对应鸡蛋的新鲜度（哈夫单位）。以它们为样本数据建立多元线性回归模型，获取鸡蛋新鲜度与其图像特征参数和声音特征参数之间的最优关系。经检验，所建模型判别鸡蛋新鲜度的正确率为92%。

Test and comprehensive model for non-destructive detection of egg freshness

In order to enhance the veracity of egg freshness non-destructive detection, a comprehensive model for non-destructive detection of egg freshness was developed, which applied image parameters reflecting egg inner color and the acoustics feature parameters acquired by knocking on eggshell. The parameters of two detection methods could be compensated and corrected each other. The lightness of egg color and coefficient of egg shape were obtained with the computer vision device. The power spectral area, formant frequency and the X center of mass were obtained by acoustics analysis. And corresponding egg freshness (Haugh unit) could be acquired through the test. Multiple linear regression model was set up by sample data based on image parameters and acoustics feature parameters. The optimal correlations between egg freshness and its image characteristic parameters and acoustics feature parameters were obtained. The test results show that the correct recognition rate of egg freshness of the model is 92%.