Expiration Identification of Canned Goods using Convolutional Neural Network

Food waste is a growing concern due to its environmental impact, aligning with Sustainable Development Goal 12.3 to halve food waste by 2030. Existing methods for checking expiration dates on canned goods can be manual and wasteful. This study proposes a system using a convolutional neural network (CNN) to identify expiration dates on canned goods. The system is built on a Raspberry Pi 4 with a camera and LCD screen. A VGG16 model is trained to recognize expiration dates in the format DD MMM YYYY using a dataset of digits and month abbreviations. The system's effectiveness is evaluated using a confusion matrix and achieves 95% accuracy on a sample of 20 different canned goods.