Image fusion method based on potential low-rank representation nested rolling guide image filtering

Abstract

The invention relates to an image fusion method based on potential low-rank representation nested rolling guide image filtering, and the method comprises the steps: carrying out the decomposition of a to-be-fused infrared image and a to-be-fused visible light image through potential low-rank representation, and obtaining a corresponding low-rank sub-layer and a significant sub-layer; performing multi-scale decomposition on the two low-rank sub-layers, extracting a detail layer, and performing weighted merging to obtain an enhanced layer; fusing the two low-rank sub-layers by using a weighted guided image filtering algorithm based on improved visual saliency mapping to obtain a low-rank sub-layer fused image; fusing the saliency sub-layers by using a pyramid decomposition-based region energy feature adaptive weighted fusion method to obtain a saliency sub-layer fused image; and adding and reconstructing the enhanced image layer, the low-rank sub-layer fusion image and the saliency sub-layer fusion image to obtain a final fusion image. According to the method, rich detail information of the source image can be reserved, the definition and contrast of the fused image are improved, and the fusion performance is good.

Description

technical field [0001] The present invention relates to the technical field of image fusion, in particular to an image fusion method and system based on latent low-rank representation nested rolling guided image filtering. Background technique [0002] Because a single image sensor cannot obtain more comprehensive and rich scene information, there are certain limitations. Therefore, people tend to use multiple sensors to capture images for image fusion, that is, to extract and integrate multiple images from the same scene at the same time through a certain algorithm. Multi-directional fusion of effective information to obtain good visual effects and rich details. Therefore, image fusion has set off a research boom in military, computer vision, remote sensing detection, medical imaging and other fields, and has very extensive and important application value. Among them, the fusion of infrared and visible light images is the basis of research in the field of image fusion, so ...

AI Technical Summary

Problems solved by technology

[0009] In order to solve the problems that the existing infrared image and visible light image fusion methods have general visual effect of the fused image, low contrast, insufficient definition, and insufficient detail texture richness, etc., the present invention provides a nesting method based on latent low-rank representation Image Fusion Method Based on Scroll-Guided Image Filtering

Images