Study of active food processing technology using computer vision and AI in coffee roasting.

Area of Science:

• Food Science and Technology
• Artificial Intelligence in Manufacturing
• Computer Vision Applications

Background:

• Modern food processing demands advanced quality control methods.
• Integrating artificial intelligence (AI) and computer vision offers solutions for complex processing environments.
• Real-time monitoring is crucial for optimizing food production.

Purpose of the Study:

• To develop and evaluate an AI-based computer vision system for coffee bean classification during roasting.
• To assess the performance of a Machine Learning (ML) model in identifying coffee beans amidst obstacles.
• To enable quantitative analysis of coffee bean color changes during the roasting process.

Main Methods:

• Development of a Coffee Bean Classification Model (CBCM) using Machine Learning (ML).
• Implementation of a computer vision system integrated with Deep Learning (DL) technology.
• Testing the model's accuracy and loss in a simulated rotating roasting machine environment.

Main Results:

• The CBCM achieved a maximum validation accuracy of 98.44% and minimum validation loss of 5.40%.
• On a test dataset, the CBCM demonstrated high accuracy (99.27%) and low loss (2.82%).
• The system successfully quantified color variations in coffee beans during roasting, even with obstacles present.

Conclusions:

• The developed AI-driven computer vision solution effectively classifies coffee beans in complex roasting environments.
• The CBCM shows high accuracy and reliability for quality control in coffee processing.
• This technology enables precise monitoring and optimization of the coffee roasting process.