Abstract: The accurate grading of mangoes based on external features is a critical task in the agricultural industry, impacting both market value and consumer satisfaction. Traditional methods of grading are labor-intensive and often subjective, leading to inconsistencies and inefficiencies. This study proposes an automated mango grading system utilizing deep learning techniques, specifically leveraging the MobileNet architecture, to address these challenges. MobileNet, known for its efficiency in mobile and embedded vision applications, is employed to classify mangoes based on visual attributes such as size, color, and surface defects. The proposed model is trained on a comprehensive dataset of mango images, annotated with corresponding grades. Performance metrics, including accuracy, precision, recall, and F1-score, are used to evaluate the system. Experimental results demonstrate that the MobileNet based grading system achieves high accuracy and robustness, significantly outperforming traditional methods. This approach promisesto enhance the efficiency and consistency of mango grading, providing a scalable solution for real-time applications in the agricultural sector.
Index Terms: Deep Learning, MobileNet, Transfer Learning, Feature Extraction, Image Classification
|
DOI:
10.17148/IJARCCE.2025.14430
