Infrared and visible image fusion based on saliency and fast guided filtering

In this study, we provide a high-quality infrared image and visible image fusion technique based on two-scale decomposition and saliency weight fusion strategy. Firstly, we propose a two-scale decomposition method based on a fast guided filter. It outperforms the traditional two-scale decomposition method in terms of computational speed and decomposition accuracy. Secondly, when we get the weight matrix of the detail layers, we use the detail layers of the two-scale decomposition as the guide image of the fast guided filter to increase the fusion weight of the detail layers, so that the fused image has rich details. Finally, the final fused image is obtained using a fast guided filter, and the brightness and contrast of the fused image are improved while retaining the fused image information. The experimental results demonstrate the fused image obtained by this method not only highlights the infrared target, but also effectively preserves the background details of the image. It outperforms the commonly used fusion algorithms in both qualitative and quantitative evaluations.