Entropy-Based Image Fusion with Joint Sparse Representation and Rolling Guidance Filter.

Area of Science:

• Computer Vision
• Image Processing
• Signal Processing

Background:

• Image fusion is crucial for applications in medicine, remote sensing, and surveillance.
• Existing methods may not fully preserve image details or achieve optimal fusion quality.

Purpose of the Study:

• To develop an advanced image fusion method utilizing multi-scale decomposition and joint sparse representation.
• To improve the performance of image fusion in terms of visual perception and objective quantification.

Main Methods:

• Joint sparse representation decomposes source images into common and innovation components.
• Weight maps are generated using joint bilateral filtering.
• Multi-scale decomposition of innovation images is performed with a rolling guide filter.
• Fused innovation and common images are combined for the final output.

Main Results:

• The proposed method achieved average metrics: Mutual Information (MI) -5.3377, Feature Mutual Information (FMI) -0.5600, Normalized Weighted Edge Preservation Value (QAB/F) -0.6978, and Nonlinear Correlation Information Entropy (NCIE) -0.8226.
• Demonstrated superior performance over state-of-the-art methods in both visual and quantitative evaluations.

Conclusions:

• The developed image fusion technique effectively integrates image information.
• The method shows significant improvements in preserving details and enhancing overall image quality compared to existing approaches.