Corn Disease Detection Using Transfer Learning

Authors : Cevher Özden

Pages : 387-393

Doi:10.34248/bsengineering.1322907

View : 80 | Download : 97

Publication Date : 2023-10-15

Article Type : Research Paper

Abstract :Detecting plant disease is a complicated yet important task to enable sustainable production in agriculture. Especially, early and on-field disease detection provides an opportunity to producers to take necessary precautions before it causes dramatic losses. Corn is one of the most important agricultural products for many countries around the world. It constitutes the main nutrient intake for large populations. This study examines and analyzes the applicability of the pretrained models in corn disease detection. A number of well-known pretrained models including Xception, ResNet50, VGG16, EfficientNetB0, MobileNet and InceptionV3 have been employed for this purpose. SMOTE is employed to solve the imbalanced data and resulting bias problem, which is a common problem in plant disease dataset. The study results indicate that SMOTE provides a good solution to the imbalanced data problem and MobileNet, VGG16 and Xception can be used as base models to develop AI applications to detect corn diseases.
Keywords : Disease detection, Classification, Transfer learning, Imbalanced data