Identification of Some Sunflower Diseases Using Deep Convolutional Neural Networks

Authors : Hakkı Fırat Altınbılek, Ünal Kızıl

Pages : 11-19

Doi:10.33202/comuagri.1387580

View : 46 | Download : 54

Publication Date : 2024-07-22

Article Type : Research Paper

Abstract :Among the oilseed plants cultivated in Türkiye, sunflower ranks first in terms of cultivation area and production. Therefore, short time detection of sunflower diseases will help producers to take necessary actions on time. Computer-based deep learning techniques have made it possible to predict these diseases with high accuracy. In this study, Google Collaboratory (GC), a free cloud-based Python coding environment, was used to detect 3 different sunflower diseases. A total of 760 images were obtained and examined in the 2022-2023 production seasons in İpsala district of Edirne province. A series of data pre-processing techniques were applied to the developed Convolutional Neural Network (CNN) model and 3 different sunflower disease prediction systems were created. It has been revealed that the model can classify with an accuracy of 0.90.
Keywords : Görüntü sınıflandırma, Derin öğrenme, Makine öğrenmesi, Ayçiçek hastalıkları