An infrared-visible light combined defogging fusion method based on collaborative learning
The infrared-visible light joint dehazing fusion method based on collaborative learning solves the problem of lost texture details in visible light images under haze by utilizing infrared-guided untangling and physical constraints. It achieves high-quality image dehazing and fusion, and is suitable for reliable perception under adverse weather conditions such as autonomous driving and intelligent monitoring.
CN122415387APending Publication Date: 2026-07-17HINTON SPACE-TIME INTELLIGENT INNOVATION RESEARCH INSTITUTE MINHANG DISTRICT SHANGHAI +1
View PDF 0 Cites 0 Cited by
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- HINTON SPACE-TIME INTELLIGENT INNOVATION RESEARCH INSTITUTE MINHANG DISTRICT SHANGHAI
- Filing Date
- 2026-05-25
- Publication Date
- 2026-07-17
Smart Images
Figure CN122415387A_ABST
Abstract
本发明提供了一种基于协同学习的红外‑可见光联合去雾融合方法，适用于在雾霾等恶劣天气条件下提升红外与可见光图像融合质量，该方法包括：构建红外引导的退化特征解缠策略，利用红外对雾霾不敏感的特性，解耦可见光图像中的雾霾退化特征；引入对比学习机制，增强多模态特征在潜在空间中的分离性，确保共有特征与模态特有特征的有效区分；基于大气散射模型构建物理约束，提升模型对雾霾退化的校正能力；在统一网络中同步实现去雾与融合，生成无雾的高质量融合图像。本发明方法克服了传统级联方法中的误差传播问题，能够在雾霾条件下有效恢复可见光纹理细节，提升融合图像的质量及其在高级视觉任务中的性能。
Need to check novelty before this filing date? Find Prior Art