December 21, 2024 | Knowledge-Based Systems |
Researchers from Universiti Malaya and Southern University College, Malaysia, conducted a study focused on addressing cognitive challenges in mangosteen production and processing by developing an automated grading system. Mangosteen is a high-value tropical fruit where quality assessment typically depends on manual inspection, which can be subjective and labor-intensive.
The study involved the construction of a specialized hardware unit designed to evaluate mangosteen fruits based on their external characteristics. Image preprocessing techniques were applied to extract color attributes using the RGB model, detect contours through image segmentation, and measure fruit diameter using ellipse fitting methods. A novel convolutional neural network (CNN) model, named MobileSeNet, was proposed and evaluated against MobileNetV2.
The MobileSeNet model demonstrated improved performance in terms of Precision, Recall, and F1-score, achieving a grading accuracy of 98.13%. It also processed images with an average speed of 76 milliseconds, which was 48 milliseconds faster than MobileNetV2. Additionally, a human-computer interaction interface was developed to support ease of use and reduce cognitive load during operation.
The findings of this study provide a foundation for further advancements in automated fruit grading systems. The approach has the potential to enhance efficiency and accuracy in mangosteen quality assessment, contributing to productivity improvements in agricultural processing.
