Year: 2025
Type: article
Abstract: This study presents a refined version of the YOLOv9 model, enhanced with the Inverted Residual Mobile Block (iRMB), to improve plant disease detection. TheiRMB module optimizes the model’s efficiency and accuracy, enabling it to effectively identify plant diseases in complex images. The integration of iRMB combines the lightweight char- acteristics of Convolutional Neural Networks (CNNs) with the dynamic processing capabilities of Transformer mod- els, enhancing feature extraction and real-time detection performance. Our YOLOv9-iRMB model was evaluated on a dataset of 2,598 images, showing significant improvement in accuracy compared to other models. It achieved a mean Average Precision (mAP) of 0.596 at an IoU threshold of 0.5 and a mAP of 0.481 averaged across IoU thresholds from 0.5 to 0.95. These results demonstrate the model’s strong performance in detecting plant diseases and its effectiveness in extracting relevant features for accurate, real-time detection
Author: Jiarui Wu
Institution: Xi'an Jiaotong University
Language: English
Topic: Smart Agriculture and AI
Subfield: Plant Science
Domain: Life Sciences
SDG: Zero hunger
Open Access status: diamond
OpenAlex ID: W4409238760