Enhanced Image Fusion through Multi-Scale Adaptive Weighting and Post-Fusion Optimization

Image fusion plays a vital role in modern image processing by integrating complementary information from multiple source images into a single, enriched representation. This capability is critical in fields such as medical imaging, remote sensing, and surveillance. However, traditional fusion methods—such as pixel averaging and wavelet-based techniques—often struggle to preserve fine details or adapt to varied image content, leading to artifacts and degraded quality. Deep learning-based approaches offer improvements but require extensive datasets and high computational resources, limiting their use in real-time or resource-constrained environments. To address these limitations, this paper proposes a novel image fusion framework combining multi-scale adaptive weighting with post-fusion enhancement. The method utilizes multi- resolution decomposition to extract frequency components, assigning perceptual-based adaptive weights based on local salience and structural relevance. A dedicated enhancement stage further improves contrast, sharpness, and detail retention. Experimental results across diverse datasets show that the pro- posed method outperforms conventional techniques, achieving higher mutual information (2.85), structural similarity (0.92), and PSNR (34.6 dB), while maintaining superior visual quality. This framework provides an efficient and robust solution suitable for real-world deployment, advancing the state-of-the-art in image fusion.