Advancing food safety through AI-driven speckle pattern imaging for milk adulteration detection

Milk adulteration is one of the most common forms of food fraud worldwide. Cow milk is frequently adulterated with water-glucose mixtures to increase volume and economic profit from vending, while goat milk, valued for its premium price and nutritional benefits, is often adulterated through undeclared addition of cheaper cow milk. Conventional milk testing methods are expensive, time-consuming, and require specialized laboratory equipment. To overcome these limitations, we present a low-cost, portable platform combining speckle pattern (SP) imaging with machine learning for rapid milk quality assessment. A laser diode illuminates the sample, generating SP images captured by a CMOS camera. Statistical features extracted from these patterns are used to train classification algorithms. Our sensing platform achieved 80% accuracy in detecting cow milk adulteration with water and glucose at different concentrations, and approximately 97% accuracy in identifying cow milk contamination in goat milk. This contactless, label-free method offers a promising tool for on-site milk authentication, enhancing food safety and supply chain integrity.