An Image Generation Algorithm Based on Illumination Attention Mechanism

The safe operation of transmission lines is under serious threat in extreme weather conditions, and visual monitoring technology based on deep learning models is an effective solution. However, existing data for transmission line scenarios under extreme weather conditions is severely insufficient. To address these issues, we propose an image generation algorithm based on the illumination attention mechanism, which includes a joint training U-Net network and an attention mechanism. Experimental results show that the proposed algorithm effectively enhances the diversity of weather recognition datasets for transmission line scenarios.