Infrared and visible image fusion method and device based on frequency domain decoupling

By using a frequency domain decoupling method, an illumination weight map is generated using global illumination factors and local brightness features. Combined with multi-scale feature map weighted fusion, the modal imbalance problem between infrared and visible light images under dynamic illumination environments is solved, and high-quality image fusion under complex illumination conditions is achieved.

CN122243768APending Publication Date: 2026-06-19CENT SOUTH UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
CENT SOUTH UNIV
Filing Date
2026-05-21
Publication Date
2026-06-19

AI Technical Summary

Technical Problem

Existing infrared and visible light image fusion technologies suffer from modal imbalance under dynamic lighting conditions and lack the ability to dynamically perceive changes in ambient light, resulting in a decline in fusion quality. This is especially true at night or in adverse weather conditions where the signal is weak and noise is severe, and in strong light or backlight scenes, there is local overexposure or loss of detail.

Method used

A frequency domain decoupling-based fusion method is adopted. By extracting the global illumination factor and local brightness features of the visible light image to generate an illumination weight map, and combining it with multi-scale spatial domain and frequency domain feature maps for weighted fusion, the feature contribution of the visible light mode is dynamically adjusted to enhance the response capability to illumination changes.

Benefits of technology

It improves the robustness and accuracy of fused images, effectively preserves infrared thermal radiation targets and visible light texture details in complex lighting scenes, and enhances the quality and information integrity of fused images.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122243768A_ABST
    Figure CN122243768A_ABST
Patent Text Reader

Abstract

This application relates to an infrared and visible light image fusion method and apparatus based on frequency domain decoupling. The method extracts multi-scale spatial domain feature maps and multi-scale frequency domain feature maps of a first image, respectively. By extracting global illumination factors and local brightness features from the first image and generating an illumination weight map, it performs weighted fusion with the multi-scale spatial domain feature maps and multi-scale frequency domain feature maps. This enables dynamic sensing and response to changes in ambient illumination, achieving differentiated processing of visible light modes. This processing method allows the frequency domain information of the visible light image to be adjusted according to its reliability under different illumination conditions, thereby better utilizing the effective information of the visible light image during the fusion process. It effectively solves the technical problems of modal imbalance, lack of ambient light sensing capability, and failure to differentiate the degradation specific to visible light modes in the current stage under dynamic illumination environments.
Need to check novelty before this filing date? Find Prior Art