Assessment of Chicken Meat Quality Through Categorized Image Analysis Using Artificial Intelligence

As consumer demands for high food quality have risen, the swift and non-destructive assessment of chicken meat freshness has become crucial for public health. The present study developed an artificial-intelligence-based digital imaging approach to estimate chicken meat quality by analyzing images categorized as healthy, defrosted, or rotten. A total of seven pre-trained convolutional neural network architectures, including Xception, Inception V3, MobileNet, DenseNet121, VGG16, VGG19, and a baseline convolutional neural network (CNN), were considered for the current study. Chicken meat samples were photographed under controlled lighting, and their quality categories were confirmed by expert inspections. Model performance was evaluated by overall classification accuracy. Among the tested architectures, the VGG19 network achieved the highest mean accuracy of 94.3%, outperforming compared to the baseline CNN by approximately 9%, while other networks exceeded 85% accuracy, indicating reliable recognition capability. This study confirmed that AI-based digital image analysis can accurately classify chicken meat freshness in a non-destructive manner. Among the tested models, VGG19 achieved the highest performance and demonstrated its strong capability for reliable meat quality assessment.