A Potential Method for Identifying Milk Adulteration and Pb(II) Contamination Scenarios Using Principal Component Analysis from Smartphone Photographs

Abstract

Heavy metal contaminants and adulteration in cow milk products are major issues affecting milk safety and quality, posing health risks to consumers of all ages. These contaminants are sometimes difficult to detect with the naked eye and can potentially pass sensory tests, particularly in white cow milk. This research explores the detection of lead(II) poisoning in milk post-production and the adulteration of different milk samples using an alternative approach through chemometric techniques based on RGB and Grey Area image analysis. A controlled photography environment was used. We analyzed over 105 samples of control, adulterated, and lead(II)-added milk in this study using image processing software. Each photograph was analyzed to provide triplicate Regions of Interest (ROI), resulting in a total of 315 statistical datasets. We found that Principal Component Analysis (PCA) effectively clustered control white milk and Pb(II)-contaminated milk. Clusters of different adulterants were recognized simply by feeding RGB and Grey Area data into PCA. However, some clusters, such as mixed chocolate milk and white milk with lead(II) contamination, were not well distinguished. In this early-stage method, a comparison study with infrared spectra will be required in future research. This alternative method shows potential promise for deployment in limited settings for real-world food quality surveillance, regulation, and biochemistry experiments.