DAONet-YOLOv8: An Occlusion-Aware Dual-Attention Network for Tea Leaf Pest and Disease Detection

2025-11-28

AI Summary

To address severe false negatives and false positives in tea leaf pest and disease detection under complex real-world tea garden conditions—including cluttered backgrounds, variable illumination, and dense foliage occlusion—this paper proposes an enhanced YOLOv8 model. Methodologically: (1) a dual-attention fusion module is introduced to strengthen feature responses from lesion regions; (2) an occlusion-aware detection head is designed to compensate for feature loss in heavily occluded targets; and (3) a C2f-DSConv module employs dynamic multi-kernel convolution to precisely model irregular lesion boundaries. Evaluated on a custom real-world tea garden dataset, the model achieves 92.97% precision, 92.80% recall, and mAP@50 of 97.10%, significantly outperforming the YOLOv8n baseline. Moreover, it reduces parameter count by 16.7%, achieving both superior accuracy and lightweight efficiency.

Technology Category

• Computer Vision: Object Detection & Categorization
• Intelligent Robots: State Estimation
• Machine Learning: Calibration & Uncertainty Quantification

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Data and user privacy-enhancing technologies for the Web
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Accurate detection of tea leaf pests and diseases in real plantations remains challenging due to complex backgrounds, variable illumination, and frequent occlusions among dense branches and leaves. Existing detectors often suffer from missed detections and false positives in such scenarios. To address these issues, we propose DAONet-YOLOv8, an enhanced YOLOv8 variant with three key improvements: (1) a Dual-Attention Fusion Module (DAFM) that combines convolutional local feature extraction with self-attention based global context modeling to focus on subtle lesion regions while suppressing background noise; (2) an occlusion-aware detection head (Detect-OAHead) that learns the relationship between visible and occluded parts to compensate for missing lesion features; and (3) a C2f-DSConv module employing dynamic synthesis convolutions with multiple kernel shapes to better capture irregular lesion boundaries. Experiments on our real-world tea plantation dataset containing six pest and disease categories demonstrate that DAONet-YOLOv8 achieves 92.97% precision, 92.80% recall, 97.10% mAP@50 and 76.90% mAP@50:95, outperforming the YOLOv8n baseline by 2.34, 4.68, 1.40 and 1.80 percentage points respectively, while reducing parameters by 16.7%. Comparative experiments further confirm that DAONet-YOLOv8 achieves superior performance over mainstream detection models.

Problem

• Detecting tea leaf pests and diseases in complex plantation environments
• Addressing missed detections due to occlusions and background noise
• Improving detection accuracy for irregular lesion boundaries

Innovation

• Dual-attention fusion module combines local and global features
• Occlusion-aware detection head compensates for missing lesion features
• Dynamic synthesis convolutions capture irregular lesion boundaries