Hyperspectral Imaging for Assessing the Quality Attributes of Cured Pork Loin

摘要

Meat is one of the most widely consumed products in the world and its market value is strongly correlated with its quality properties such as chemical composition, technological and sensory attributes. Therefore, food industry needs to minimize the variability of meat quality properties in order to keep the quality standards as high as possible. The current industrial methods for assessing the meat quality properties are destructive, time consuming and consequently unsuitable for the on-line application classification of meat quality properties. On the other hand Hyperspectral Imaging (HSI) allows on-line high-throughput screening in a non-destructive nature. We have use HSI with random forest classification on cured pork loin meat for classify their quality standards with a classification accuracy of 95%.