Enhanced Classification of Meat Freshness Using Lightweight Neural Networks

Abstract

The freshness of food, particularly meat, significantly impacts the flavour of dishes served in restaurants and the health of consumers. However, in the absence of records indicating the production date of meat, the current method of determining meat freshness primarily relies on human subjective judgment. This method is heavily dependent on the individual's experience, resulting in considerable uncertainty. To address this issue, this study proposes an innovative approach that uses artificial intelligence and non-invasive methods to distinguish the freshness of meat based on photographs. Experimental results show that the proposed method achieves an accuracy of over 98% in distinguishing meat freshness.