Comparison of milk-clotting activity measurement using the Berridge's operator-based approach and a proposed digital method

Abstract

The milk clotting activity (MCA) analysis of coagulants used in the production of cheese is an important parameter used in the modern dairy industry. The MCA is calculated using the estimated milk-clotting time (MCT). The standard method used for estimation of MCT is presented in the ISO 23058/IDF 199 Standard using the Berridge's operator-based approach. In this paper a new method is presented to estimate the MCT using digital image acquisition and signal processing. The estimate is obtained by processing a sequence of images of the wall of a round bottom flask containing milk and the coagulant. Digital signal processing and filter techniques were used to remove interferences and artifacts and improve the clotting time estimation. A set of 28 samples of milk plus coagulant were processed. The method's accuracy is about 1/30 s and the average relative error, when compared to an experienced human operator, was με = 2.82%, according to the experimental results, with an associated standard deviation of σε = 3.07% and a median of mε = 1.92%. The method may be used to replace the human operator during the MCT estimation procedure.