基于跨层协同交互与空频联合感知的遥感图像去雾OA

In view of the image degradation,texture loss,and color distortion in the dehazing process of the existing U-shaped dehazing networks,this paper proposes a remote sensing image dehazing network based on cross-layer collaborative interaction and spatial-frequency joint perception.Specifically,a cross-layer collaborative interaction module is introduced.This module can effectively capture long-range dependencies among cross-layer features,expand the local receptive field,and facilitate the collaborative interaction of global and local information during both encoding and decoding stages,so as to enhance the dehazing performance.Additionally,a spatial-frequency joint perception module is proposed.This module can realize the complement of the spatial domain and frequency domain by fusing spatial and frequency features,so as to improve the recovery of image texture and color.Experiments show that the proposed algorithm outperforms the existing methods on the public remote sensing haze datasets HRSD and RRSHID-Thick.To sum up,this method has a certain applicability.