Visualized detection of quality change of cooked beef with condiments by hyperspectral imaging technique.

The heat treatment and seasoning of meat are indispensable before its consumption. In this work, the spectral characteristics of cooked meat and condiments were analysed by hyperspectral imaging (HSI) technology. The spectral reflectance of spices was significantly lower than that of meat protein, and that the spectral reflectance of protein regularly increased upon heating at 800-956 nm range. PCA pre-process and SVM models were used to predict beef moisture (R (2) = 0.912) and tenderness (R (2) = 0.771) based on 100 beef data. Mapping technology clearly showed the dynamic change of meat tenderness during heating, and the performance of 3D mapping was better than that of 2D mapping. Based on 750 nm/900 nm ratio image and machine-vision method, spice uniformity was accurately calculated. Thus, the quality of cooked meat and condiments distribution can be simultaneously evaluated by HSI. This technology can be used in the intelligent production of complex meat products in the future.