A Point-Line-Based SLAM Framework for UAV Close Proximity Transmission Tower Inspection

Abstract

Employing unmanned aerial vehicles (UAV) to conduct close proximity inspection of transmission tower is becoming increasingly common. This paper aims to solve the key problem of close proximity navigation - estimating the UAV positions. To this end, we propose a novel monocular-vision-based environmental perception approach and implement it in a hierarchical embedded UAV system. The proposed framework is an improved point-line-based SLAM framework consisting of feature matching, frame tracking, local mapping, loop closure, and nonlinear optimization. In order to enhance frame association, the prominent line feature of tower is heuristically extracted and matched followed by the intersections of lines are processed as point feature. Then the bundle adjustment optimization leverages the intersections of lines and the point-to-line distance to improve the accuracy of UAV localization. Additionally, two reasonable paths are planned for the refined inspection. In experiments, the whole UAV system developed on Robot Operating System (ROS) network is evaluated along the paths in a real-world inspection environment, which achieves a satisfactory result.