Visual navigation and obstacle detection algorithm of power inspection UAV

In the visual navigation and obstacle detection of power inspection UAV, how to identify and avoid obstacles efficiently and accurately has become a key technical problem. This paper presents A new method, Single Shot MultiBox Star (SSMS), which combines A Single Shot MultiBox Detector (SSD) and A* (A-STAR) path planning algorithm. SSMS method realizes fast and accurate target detection through SSD technology, and uses A* algorithm to plan the optimal obstacle avoidance path, so as to realize efficient navigation and obstacle detection of power inspection UAV. Finally, through the comparison experiment and analysis between the architecture of this paper and the traditional detection algorithm, the experimental results show that the SSMS method has significant advantages in improving the detection accuracy and enhancing the robustness of the system, and provides a strong technical support for the intelligent navigation of the power inspection UAV. This proves the feasibility and effectiveness of this model in visual navigation and obstacle detection of power inspection UAV.