In this research, a nondestructive testing method based on knock‐generated vibration acoustic signals was explored for measuring the content of total sugar in sweet potato for the first time. Vibration acoustic signals were acquired using a self‐made signal collection platform and processed on the basis of time‐frequency theory and Hilbert transform. Results indicated that total sugar content showed significantly positive correlations with waveform index, sound intensity and frequency energy of 600–800 KHz (p < .01), but negative correlations with frequency energy of 0–1 KHz, 3–4 KHz, 5–6 KHz, 9–10 KHz, and 0–200 Hz (p < .01). The eigenvalues could be reduced to three principle components by principle component analysis. Hence, vibration acoustical signals could be recognized as an ideal nondestructive method for predicting total sugar concentration in sweet potato.
Practical applications
Although the chemical methods are commonly used for quantitative analysis of food components, they suffer from some disadvantages, such as time‐consuming procedures and destructiveness of tested samples. In recent 30 years, acoustic technology has been widely applied in food quality classification and detection, especially vibration acoustical signals are considered as alternative technology for measuring the texture and quality of food. The acoustical signals of knocking vibration belong to mutation signals, which are frequently excavated and analyzed by Hilbert‐Huang transform. Thus, in this work, a relationship between total sugar of sweet potato and acoustic eigenvalues mainly analyzed by Hilbert‐Huang transform was established for the first time. It is hoped that present study will provide evidence for the applications of acoustic signals on foods detection without sample destructiveness.