多旋翼无人机单目V-SLAM研究综述

The visual simultaneous localization and mapping (VSLAM) is an effective approach to solve the problem encountered in the works of multi-rotor unmanned aerial vehicle (UAV) under unknown environment, which is very useful for the location, navigation and path planning of multi-rotor UAV. The main research achievements and the latest progresses of monocular V-SLAM for multi-rotor UAV are collected. Firstly, the classification and history of V-SLAM are briefly described, the principle and characteristics of several representative monocular V-SLAM methods based on filtering, optimization, direct method and IMU fusion are summarized. Secondly, the feasibilities of monocular V-SLAM for multi-rotor UAV are analyzed, and the overseas and domestic research teams of monocular VSLAM for multi-rotor UAV are also summarized. Finally, several promising research directions for further investigation are discussed.