Multi-scale detection method of power lines suitable for complex scenarios

Transmission lines, as an important component of the power system, play a crucial role in delivering electrical energy from power plants to users. With the continuous expansion of the power system and technological advancements, the accuracy and efficiency of transmission line detection have become increasingly important. Traditional image processing methods have achieved some success in simpler backgrounds, but in complex backgrounds or low signal-to-noise ratio (SNR) environments, they face challenges such as low accuracy and poor anti-interference performance. To address these issues, this paper proposes a transmission line detection algorithm based on feature keypoints. Instead of using traditional pixel-based segmentation, this method employs keypoints to represent the geometric features of transmission lines. By considering the local structure and geometric features of the target, the method effectively improves detection accuracy. To further enhance performance in low SNR environments, a global dual-attention mechanism is introduced. The spatial attention mechanism focuses on key regions in the image, while the channel attention mechanism adaptively adjusts the importance of different features, thus enhancing the model's feature extraction capability. This is especially effective in detecting long, thin transmission lines. Experimental results demonstrate that the proposed algorithm outperforms traditional methods in complex environments and can detect transmission lines more accurately, offering significant practical value and innovation.