Predicting the sensory crispness of coated turkey breast by its acoustic signature

Crispness of coated turkey breast after frying, and reheating by microwave or oven, was evaluated by sensory, mechanical tests and a newly developed acoustic signature-classifying model (ASCM). The acoustic signatures were recorded during compression-tests and analyzed by the ASCM. Support vector machines (SVM) and Fuzzy KNN classifiers were used to performed classification of the acoustic signals to frying, and heating by microwave or oven. The sensory crispness scores and the initial slope of the force deformation curve were predicted using frequency domain spectra of the acoustic signals by neural networks (NN). The results showed that a linear classifier (LC) provides satisfying predictions of the sensory crispness grades and the initial slope of the force deformation curve for fried and microwave heated samples.