Smartphone determination of fat in cured meat products

A method has been developed to determine the fat content in different cold meat products by image processing using the camera of a mobile phone. Salchichón, chorizo, salami and cured ham pictures were taken with a Meizu M2 Mini mobile phone camera under fixed lighting conditions of the light emitting diode flash of the mobile phone. Images were treated with Matlab to obtain the mean pixels of average red, green and blue camera values colours (RGB) of the pixels and different data pretreatments were taken into account to correlate colour parameters with fat content values determined in a series of commercially available samples by the Soxhlet method. RGB values were used as input variable and its correlation with fat content of samples was studied using Partial Least Squares (PLS) and Support Vector Machine (SVM). The best correlation between fat content and RGB colour descriptors was found in salchichón and salami samples using SVM, with relative errors of calibration, cross-validation and prediction of 18%, 20% and 16%, respectively. So, the use of a Smartphone camera provides an easy, low-cost, eco-friendly and rapid method for fat content determination in cold meat products.