A Review on Chilli Disease Detection Using Deep Learning

Abstract

Agriculture is one of the most important sectors in the world, especially in India, where a large proportion of the population depends on farming for their livelihood. Chilli (Capsicum annuum) is a widely grown crop used extensively in food, medicines, and spices. However, chilli plants are frequently affected by diseases such as leaf curl, bacterial leaf spot, powdery mildew, mosaic virus, and anthracnose, which significantly reduce crop quality and yield, causing major economic losses to farmers. Traditionally, disease identification is performed manually by farmers or agricultural experts, a process that is time-consuming, costly, and prone to error, particularly across large farming areas. With the rapid growth of Artificial Intelligence (AI) and Deep Learning (DL), smart agriculture systems capable of automatically detecting diseases from leaf images have become increasingly practical. This paper reviews and compares several deep learning architectures, including Convolutional Neural Networks (CNN), YOLO, ResNet, MobileNet, and Vision Transformers (ViT), for chilli disease detection, evaluating their accuracy, precision, recall, and computational efficiency. The role of image preprocessing, feature extraction, and data augmentation in improving model performance is also discussed, along with key challenges such as limited datasets, variable environmental conditions, and high computational cost. Finally, the paper outlines future research directions including IoT-based monitoring, drone-assisted imaging, and explainable AI for sustainable precision farming.