Plant field pest fine-grained recognition method, system and device based on deep learning and storage medium
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- ZHEJIANG UNIV
- Filing Date
- 2025-07-17
- Publication Date
- 2026-04-21
AI Technical Summary
Existing pest identification methods are difficult to classify pests with similar morphologies accurately in the field, resulting in low identification precision and being time-consuming and labor-intensive, which cannot meet the needs of agricultural production.
A deep learning-based approach is adopted, using the MaxViT model combined with a deformable self-attention mechanism and a contrastive learning loss function to train a plant field pest identification model. End-to-end fine-grained classification is performed through convolutional modules, MaxViT-DAT attention modules, global average pooling modules, and fully connected layer modules.
It enables fine-grained classification of various morphologically similar pests with an identification accuracy of over 90%, improving the efficiency and precision of pest identification and making it suitable for agricultural pest surveys and insect ecological surveys.
Smart Images

Figure CN120913167B_ABST
Abstract
Citation Information
Patent Citations
Unsupervised echocardiogram section identification method
CN115578589A