MTDL-EPDCLD: A Multi-Task Deep-Learning-Based System for Enhanced Precision Detection and Diagnosis of Corn Leaf...

Area of Science:

• Agricultural Science
• Computer Science
• Artificial Intelligence

Background:

• Corn leaf diseases cause significant agricultural losses and threaten global food security.
• Accurate and timely disease detection is vital for effective crop management.

Purpose of the Study:

• To develop a multi-task deep learning system (MTDL-EPDCLD) for enhanced precision detection and diagnosis of corn leaf diseases.
• To create a mobile application for accessible disease identification.

Main Methods:

• A two-task system: Task 1 (RAHSI) uses a CNN-4 model with spatial attention for health status identification.
• Task 2 (FDCA) employs a customized MobileNetV3Large-Attention model for fine-grained disease classification.
• The system was developed using the Qt framework for cross-platform compatibility.

Main Results:

• Task 1 achieved 98.73% accuracy in distinguishing healthy from diseased corn leaves.
• Task 2 demonstrated a validation accuracy of 94.44%, with 4-8% improvements in precision, recall, and F1 score over other models.
• The model achieved an AUC of 0.9993, outperforming mainstream models.

Conclusions:

• The MTDL-EPDCLD system offers an accurate and efficient solution for corn leaf disease detection and diagnosis.
• This technology supports informed disease management decisions, potentially increasing crop yields and enhancing food security.
• The research presents a promising advancement for agricultural practices in disease management.