MGFusion: A New Method for Infrared and Visible Image Fusion
Author Information
Author(s): Yang Zengyi, Li Yunping, Tang Xin, Xie MingHong
Primary Institution: Kunming University of Science and Technology
Hypothesis
Can a multimodal large language model improve the quality of infrared and visible image fusion?
Conclusion
The proposed method significantly enhances the quality of fused images by leveraging semantic information from a multimodal large language model.
Supporting Evidence
- The proposed method outperforms existing methods in both visual quality and objective evaluation metrics.
- Experimental results validate the effectiveness and superiority of the proposed method on multiple public datasets.
Takeaway
This study shows that using a special model can help combine infrared and visible images better, making them clearer and more useful.
Methodology
The study employs a multimodal large language model to enhance image features and improve fusion quality through a new framework.
Limitations
The method may not perform as well on other types of image fusion tasks without retraining.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website