一种基于深度学习的车牌识别方法及系统
By using the YOLOv8 algorithm and the improved LPRNet network model, combined with cross-attention and multi-scale feature fusion modules, the problems of high complexity of license plate recognition models and inaccurate recognition in complex scenarios are solved, and efficient license plate recognition under different lighting and weather conditions is achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- HUARONG TECH CO LTD
- Filing Date
- 2025-04-30
- Publication Date
- 2026-07-17
AI Technical Summary
Existing license plate recognition technology models are highly complex and lack accuracy and reliability in complex scenarios, especially under different lighting conditions and complex weather conditions.
By employing the YOLOv8 algorithm model and the improved LPRNet network model, combined with the cross-attention module and the multi-scale feature fusion module, the robustness and accuracy of the model are improved through license plate target detection, segmentation and character recognition.
It effectively reduces the complexity of license plate recognition models, improves the reliability and accuracy of recognition in complex scenarios, and enhances the model's recognition capabilities under different lighting and weather conditions.
Smart Images
Figure CN120496046B_ABST