基于DSP的鸡蛋破损检测方法

Experimental Study on DSP System of Eggshell Crack Detection

通过分析受敲击鸡蛋的声音功率谱图而确定出特征参数:共振峰对应的模拟量频率值、功率谱面积、高频段额外峰功率谱幅值和第32点前后频段功率谱面积的比值,并通过Bayes理论建立鸡蛋破损判别模型。然后采用DSP鸡蛋破损检测系统,进行鸡蛋破损检测的试验。检测结果表明,此系统判别精度高,方法可行。

Characteristic parameters such as the peak resonant frequency, the values oi the power spectrum, the power spectrum swing of the high frequency range and the ratio of the area of power spectrum between front and back on the 32 points in spectrum were found for differentiating cracked eggs and intact eggs according to the analysis of their power spectra in this research. Bayes theorem was applied to develop discrimination models. Finally, a testing and grading based on DSP was desifned. The test showed that the system has a good precision and feasibility.