This research develops an intelligent system for meat quality grading, improving assessment accuracy in meat processing environments, suggesting enhanced food safety.
Key Points
The central aim is to develop a system for detecting meat spoilage and grading quality accurately, addressing weaknesses in manual assessments.
Developed a new dataset for meat spoilage and quality detection.
Trained a particle swarm optimization-based convolutional neural network using the dataset.
Integrated the trained model into a Raspberry Pi 4 for stand-alone operation.
Conducted comparative analyses against a baseline CNN.
Achieved a 2.91% increase in accuracy compared to baseline CNN.
Improved precision by 2.49%, F1-score by 0.99%, recall by 1.87%, specificity by 2.74%, and sensitivity by 1.14%.
Indicated potential for enhanced food safety in processing and retail environments.