The potential of acoustic signals of rolling eggs on an inclined plate and impulse response for nondestructive detection of eggshell crack was investigated. Discriminations of hairline cracked and star cracked eggs from intact ones were carried out using artificial neural network. Ten features were used based on one-way ANOVA F-test statistics. According to the result, holdout detection accuracy of the inclined plate and impulse response methods were 92.3% and 94.6%, respectively. The results indicated that these two methods were potentially useful for discrimination of eggs according to detection of different eggshell cracks.
