Low light and fog are among the key factors that significantly impact image quality. In low-illumination conditions, due to the limited amount of light that can be captured by the UAV (Unmanned Aerial Vehicle) camera, the image becomes dark, and the hue shifts. These conditions seriously affect subsequent advanced image processing tasks for UAV images. To address this challenging problem, this paper designs a spatially aware enhanced dehazing algorithm for low-illumination scenes using an end-to-end model. The design is inspired by the concepts of spatial contextual information and attention mechanisms, effectively addressing issues such as color distortion, patchiness, and loss of edge information caused by traditional dehazing algorithms without relying on the atmospheric scattering model. Finally, the effectiveness of the proposed algorithm in low-illumination scenes is demonstrated through comparative experiments.