Recent Advances in Deep Learning-Based Disease Detection in Garlic Plants in Agriculture: A Review

Abstract

AI and DL technologies have rapidly evolved and revolutionized modern farming practices, especially in automated plant disease detection and precision farming. This crop is economically and nutritionally significant and highly susceptible to diseases that affect yield, quality of bulbs and market value, including white rot, basal rot, purple blotch, downy mildew, rust and viral infections. Diagnosing disease promptly and correctly is very important for production losses and sustainable garlic production. Diagnosing diseases using conventional approaches that rely on human eye observation and experience can be challenging for large-scale agriculture due to their potential drawbacks, such as being costly, subjective, labor-intensive, and time-consuming. Automatic disease identification by image analysis using DL approaches has been more popular in the past few years. Finding diseases in garlic using Deep Learning (DL) techniques including YOLO, CNNs, Vision Transformers (ViT), hybrid CNN-ViT, and Generative Adversarial Networks (GAN) is the focus of this publication. Furthermore, the study discusses current research progress, key challenges, limitations, and future opportunities for developing accurate, scalable, and intelligent garlic disease detection systems for precision agriculture.