Multivariate accelerate shelf life test (MASLT) based on hyperspectral imaging to estimate shelf life of cape gooseberry (Physalis peruviana L.)

Abstract

For first time, it was proposed a multivariate shelf life estimation of cape gooseberry (CG) based on Multivariate Accelerating Shelf Life Testing (MASLT) and Near Infrared Hyperspectral Imaging (NIR-HSI). CG samples were stored at 6, 10 and 20 °C for 0, 3, 6, 9 and 12 days. Each day samples were imaged, evaluated by a sensory panel and characterized in physicochemical composition. The best region of interest in hyperspectral images was selected, then a Principal Component Analysis (PCA) was performed. The time-related PC2 scores were extracted, used to build kinetic charts and fit to zero-order reaction (R2 > 0.82). From the sensory evaluation, we defined cut-off value of -0.3576, corresponding to the sample stored at 20 °C for 6 days. Finally, the shelf life was estimated at 7.5 days, 15.4 days and 20.5 days for cape gooseberry stored at 20, 10 and 6 °C, respectively. In conclusion, we developed a non-contact method based on NIR-HSI to estimate the shelf life of cape gooseberry, which is related to global degradation, being more realistic.