Anisotropic Guided Filtering and Multi-level Disintegration Method for NIR and Visible Image Fusion

The fusing of near-infrared and visible images holds substantial research relevance within the domain of image fusion. This paper presents a novel approach for fusing near-infrared and visible images, utilizing prospective multi-level disintegration and Anisotropic Guided Filtering (AnisGF). The proposed process involves using a Multi-level image Disintegration method based on Latent Low-Rank Representation (MDLatLRR) to carry out the multi-layer disintegration procedure, enabling the extraction of detailed data from the source images. Additionally, it softens the image to collect global image data. AnisGF provides good edge protection, reducing halo artifacts. When MDLatLRR and AnisGF are fused, the fused image reveals near-infrared heat and visible texture. The results of our experiment demonstrate that our approach surpasses five other established methodologies in terms of qualitative and quantitative measurements.