A Review on Meat Quality Evaluation Methods Based on Non-DestructiveComputer Vision and Artificial Intelligence Technologies

摘要

Increasing meat demand in terms of both quality and quantity in conjunction withfeeding a growing population has resulted in regulatory agencies imposingstringent guidelines on meat quality and safety. Objective and accurate rapidnon-destructive detection methods and evaluation techniques based on artificialintelligence have become the research hotspot in recent years and have beenwidely applied in the meat industry. Therefore, this review surveyed the keytechnologies of non-destructive detection for meat quality, mainly includingultrasonic technology, machine (computer) vision technology, near-infraredspectroscopy technology, hyperspectral technology, Raman spectra technology, andelectronic nose/tongue. The technical characteristics and evaluation methodswere compared and analyzed; the practical applications of non-destructivedetection technologies in meat quality assessment were explored; and the currentchallenges and future research directions were discussed. The literaturepresented in this review clearly demonstrate that previous research onnon-destructive technologies are of great significance to ensureconsumers’ urgent demand for high-quality meat by promoting automatic,real-time inspection and quality control in meat production. In the near future,with ever-growing application requirements and research developments, it is atrend to integrate such systems to provide effective solutions for various grainquality evaluation applications.