Application of quality control in product supply using deep learning and machine learning
Özet
Artificial intelligence is the source technology for the now common in quality control. In this research, two novels deep learning and machine learning-based models have been proposed in order to supply quality control in vegetable and fruit supply. The purpose is to identify tomatoes, apples, oranges, tangerines and potatoes, which are among the most cultivated vegetable, fruit and root vegetable types in Turkey, as fresh or spoiled. The dataset that was used in the study was constructed based on images that were collected online as well as snapped by using a mobile phone, and it was organized to contain both fresh and decayed items. The developed methods are evaluated using accuracy, precision, sensitivity, and F1 score performance criteria. With 97.57% general accuracy, the deep learning model was very effective while the machine learning model obtained an accuracy rate of 90.20%, which is a good substitute method. The deep learning model achieved better performance as it modeled more complex features in the visual data, while the machine learning model was characterized by its rapid processing speed and low hardware requirements. The results suggest that DL techniques achieve better performance in the quality control task, whereas ML-based methods may be also preferred in some specific scenarios. This study addresses fruit and vegetable supply processes, providing quality control and being able to quickly, correctly and economically classify the products, contributing to the artificial intelligence-based quality control automation.
Anahtar Kelimeler:
Deep Learning Machine Learning Quality Control Supply Chain Managementİndirmeler
Referanslar
İndir
Yayınlanmış
Sayı
Bölüm
Nasıl Atıf Yapılır
Lisans
Telif Hakkı (c) 2026 Holistence Publications

Bu çalışma Creative Commons Attribution 4.0 International License ile lisanslanmıştır.


