Non-linear and selective fusion of cross-modal images

Existing image fusion methods pay little research attention to human visual characteristics. However, characteristics play an important role in processing tasks. To solve this problem, we propose a cross-modal method that combines illuminance factors and mechanisms. Human are studied simulated task. Firstly, order reject high low-frequency mixing reduce the halo effect, perform multi-scale decomposition. Secondly, remove highlights, saliency map deep feature combined with factor high-low frequency non-linear fusion. Thirdly, maps selected through channel network obtain final map. Finally, validate our on public datasets of infrared visible images. The experimental results demonstrate superiority under complex illumination environment. In addition, also effectiveness simulation