Olive Fruit Selection through AI Algorithms and RGB Imaging.

Area of Science:

• Agricultural Science
• Computer Science
• Food Science

Background:

• Extra virgin olive oil quality is dependent on fruit quality.
• Optical selection aids in producing high-quality oils from diverse fruit batches.
• Current methods require manual assessment of olive quality.

Purpose of the Study:

• To evaluate a convolutional neural network (CNN) algorithm for classifying olives into quality grades.
• To assess the algorithm's potential for integration into industrial sorting processes.
• To determine the sorting performance of the AI-driven classification system.

Main Methods:

• Olives were visually analyzed and categorized into five classes by trained operators based on veraison and defects.
• Images were captured using an industrial RGB camera on a conveyor belt system.
• AI techniques, specifically CNN, were employed for image analysis and classification.

Main Results:

• AI-based modeling approaches demonstrated excellent performance on RGB image datasets.
• The CNN algorithm achieved high accuracy in classifying olives into predefined quality categories.
• The study confirmed the effectiveness of automated visual analysis for olive grading.

Conclusions:

• The developed approach shows significant potential for qualitative discrimination of olive fruits.
• The system is suitable for evaluating sorting machine performance.
• This AI-driven method can be implemented in industrial sorting processes for enhanced efficiency and quality control.