An Active Vision Based Localization and Mapping Method for MAV Inspection in Visual-degraded Restricted Environment

摘要

This paper proposed a localization and mapping method based on active vision for the tasks of Micro Aerial Vehicles(MAV) inspection in visual-degraded restricted environment. In order to solve the defects of unobservable and texture-less in visual-degraded environment, the cooperative markers that provide distinct features in dark environments and the active vision sensor combining structured light and monocular vision is designed. A robust MAV inspection localization method consist of cooperative marker based pose measurement and structured light reconstruction. In addition, a hybrid restricted environment map containing the relative pose of adjacent cooperative markers and laser point cloud is constructed. Experimental results show that the proposed method effectively solves the problem of inadequate observable features in visual-degraded environment, achieves robust localization and dense map construction, which provides a effective solution for visual localization and mapping in visual-degraded restricted environment.