VISION TRANSFORMER-BASED APPROACH FOR WALNUT LEAF CLASSIFICATION
Kahramanmaraş Sütçü İmam Üniversitesi Mühendislik Bilimleri Dergisi · 03.06.2026 · 29 (2) · 608-616
DOI: 10.17780/ksujes.1805431 ↗Kurz gefasst
Fachartikel in der Zeitschrift Kahramanmaraş Sütçü İmam Üniversitesi Mühendislik Bilimleri Dergisi, online erschienen am 03.06.2026. Verfasst von Ali Alper Demir und 1 weiteren Autor:innen. Die Arbeit nennt 20 Literaturangaben.
Abstract (Originalsprache)
Walnut is widely utilized across various industrial sectors such as food, pharmaceuticals, and cosmetics due to its high nutritional value and numerous health benefits. Turkey is a major walnut producer, with many species both locally and globally. However, these species exhibit a high degree of morphological similarity in terms of leaf shape and color, making it difficult to distinguish between them through visual inspection alone. This challenge increases the likelihood of errors in manual classification processes. Consequently, there is an increasing demand for AI-based systems for accurate automatic classification of walnut leaves. This study utilizes a vision transformer (ViT) approach, which, through its attention mechanism, enables more precise extraction of visual features than conventional techniques. To evaluate the effectiveness of a ViT-based classification model, an open-source dataset was utilized, consisting of walnut leaf images collected from trees located in the experimental orchard of the Yalova Atatürk Central Horticultural Research Institute. A total of 1668 leaf images of 17 different walnut species in this dataset were used in the training and testing process of the ViT model. The model achieved 95.51% accuracy, demonstrating superior performance in classifying walnut leaf images compared to other image classification methods.
Zitieren
APA
Demir, A. A., & Demir, G. D. (2026). VISION TRANSFORMER-BASED APPROACH FOR WALNUT LEAF CLASSIFICATION. Kahramanmaraş Sütçü İmam Üniversitesi Mühendislik Bilimleri Dergisi, 29(2), 608-616. https://doi.org/10.17780/ksujes.1805431
Vancouver
Demir AA, Demir GD. VISION TRANSFORMER-BASED APPROACH FOR WALNUT LEAF CLASSIFICATION. Kahramanmaraş Sütçü İmam Üniversitesi Mühendislik Bilimleri Dergisi. 2026;29(2):608-616. doi:10.17780/ksujes.1805431
BibTeX
@article{Demir2026,
author = {Demir, Ali Alper and Demir, Gizem Dursun},
title = {VISION TRANSFORMER-BASED APPROACH FOR WALNUT LEAF CLASSIFICATION},
journal = {Kahramanmaraş Sütçü İmam Üniversitesi Mühendislik Bilimleri Dergisi},
year = {2026},
volume = {29},
number = {2},
pages = {608--616},
doi = {10.17780/ksujes.1805431}
}
EUVIMED