The Application of Machine Learning in Differentiating Broth Containing Pork Fat and Chicken Fat Using UV LED Fluorescence Imaging System

Abstract

In Indonesia, individuals have been found engaging in fraud for selling soupy dishes by adding pork fat to the broth. It is quite challenging to identify the pork fat contaminated soup from other halal broth. Using Machine learning, this studi attemps to identify and differentiating between RGB (Red Green Blue) values in picture of broth tainted with chicken and pork fat. The successful detection and differentiation of RGB values in broth contaminated with pork fat and chicken fat have been achieved. The broth samples were detected using a high-power UV-LED (Ultra Violet-Light Emitting Diode) Fluorescence Imaging System, while differentiation was accomplished through the implementation of a machine learning system. The data were processed using RapidMiner software with the K-NN algorithm. Detection was successfully performed through the spectrum of RGB values generated, while differentiation achieved a accuracy of 100%, precision of 100%, recall of 100%, and an AUC of 1.0.