Year: 2025
Type: article
Abstract: Global food security and economic stability depend heavily on agriculture. Effective and prompt detection is crucial for minimizing losses and ensuring crop health since crop diseases represent serious risks to output and quality. Conventional disease detection techniques mostly depend on manual inspection, which is laborious and prone to mistakes. Promising approaches to automating and improving disease detection procedures are provided by recent developments in deep learning (DL). Despite the availability of sophisticated algorithms and high-resolution imaging technologies, real-time crop disease detection remains challenging due to variations in lighting, plant conditions, and disease symptoms.Two primary approaches were employed: image classification as well as object detection. Algorithms tested include YOLO (You Only Look Once), Convolutional Neural Networks (CNNs), along Faster R-CNN. Our evaluation prioritized precision and recall to ensure effective disease identification. Among the models tested, YOLOv4 demonstrated the most promising results, achieving an F1 score of 88.
Author: Sangeeth Elangovan
Institution: St. Joseph’s College of Engineering
Language: English
Topic: Smart Agriculture and AI
Subfield: Plant Science
Domain: Life Sciences
SDG: Zero hunger
Open Access status: diamond
OpenAlex ID: W4408738518