Abstract:
This study sought to identify a method for swiftly improving the correlation between sensory crispness and acoustic characteristics of puffed food. The sensory group was formed by selecting individuals adept at identifying brittleness through sequencing test, with sensory assessments conducted on various puffed food. Adopting 1000 Hz as the interval for energy segmentation, the Hilbert-Huang Transform (HHT) was utilized to segment the energy of the sound signal from puffed food to observe the process of energy migration. The relationship between sensory evaluation and acoustic properties of 10 varied puffed food types was studied. The results of sensory evaluation showed that samples with higher "crispness" were more likely to obtain higher evaluation scores. The pulse factor (
r=0.937) and kurtosis (
r=0.889) showed an extremely significant correlation with sensory crispness between different types of puffed food, revealing that the energy release characteristics remained consistent (
P<0.01), while the energy released by the crunchy foods led to abrupt and unstable signals. It was observed that the sound signal frequencies of puffed food predominantly concentrated in the lower frequency range, accounting for 80% of the total energy. The results showed that there was no significant correlation between the low frequency interval and sensory crispness among different types of puffed food, but the acoustic features extracted from the inherent low frequency interval were significantly correlated with sensory crispness. This study establishes a fast method to identify the best frequency interval of sound signal recording of puffed food, thereby enhancing the accuracy and reliability of acoustic evaluation in food sensory quality control.