A portable electrochemiluminescence imaging system based on image processing for real-time detection of melamine

In this study, a portable electrochemiluminescence (ECL) imaging system based on Raspberry Pi was developed for the instant detection of melamine. The system employed a lithium battery as the excitation source and Ru(bpy)32+/TPA as the reaction system. A homemade 3D-printed black box served as a reaction chamber for the ECL reaction. ECL images were captured by a camera and subsequently classified by the MobileNet model. Over 500 images obtained from the experiment were used as a self-built dataset to train the neural network model. As the core processor, Raspberry Pi was not only used to control the camera, but also to perform Image enhancement and color analysis. The experimental results were further validated using image processing algorithms. The system demonstrated highly sensitive detection of melamine under optimal conditions, with a linear fit curve in the range of 10−8–10−3 M and a detection limit of 3.54 nM. In conclusion, this system holds great potential for the instant detection of melamine.