Efficient grasping of UAV robotic arms for power transmission line inspection based on visual servoing technology

With the expansion of power systems and increasingly complex operating environments, traditional transmission line inspection has become inefficient and high-risk, failing to meet the demands of modern power grids. This paper proposes a holographic panoramic reconstruction method for transmission lines based on UAV intelligent vision and an efficient robotic-arm grasping strategy. Multi-angle images are collected by a UAV equipped with an HD camera, IMU, and GPS, and transmitted in real time via a 4G communication network. The proposed framework includes edge and keypoint heat-map extraction, attitude correlation field construction, skeletonized reconstruction, and central-axis position estimation, while reconstruction accuracy is enhanced through an improved stereo-matching algorithm. Experimental results under various typical scenarios demonstrate that the proposed method outperforms BM, SGBM, and GC in reconstruction accuracy, edge preservation, and noise suppression, effectively supporting autonomous UAV robotic-arm grasping with strong robustness and practical value.