A REAL-TIME PORTABLE DIAGNOSTIC IMAGING DEVICE FOR FLUORESCENCE IMAGE ACQUISITION IN ARTIFICIAL INTELLIGENCE-BASED MILK QUALITY DETECTION

Abstract

This invention relates to a portable system and device called PRIME Box which is used to detect milk quality based on fluorescence biomarker visualization and artificial intelligence (AI) based classification. PRIME Box consists of several main components, including: a closed acrylic box structure measuring ±12.5 × 12.5 × 15 cm, a 405 nm violet LED light source for fluorescence excitation, a CMOS camera of at least 12 MP, and a platform for placing milk samples in a petri dish. The process of use begins with the addition of a fluorescent dye indicator (such as Methyl Red) to the milk, then imaging is performed under controlled lighting conditions. The resulting image is then sent to a Flutter-based mobile application that has been integrated with an AI model (MobileNetV2) in .tflite format for milk classification into three quality categories: Fresh, Suitable, and Damaged.This system enables rapid, non-destructive detection of milk spoilage in real-time in the field, without the need for complex laboratory testing. The primary objective of this invention is to provide a portable, efficient milk quality acquisition and classification system that can be directly used by farmers or cooperatives as a tool for rapid and objective milk quality detection.