Abstract:
Unmanned aerial vehicle (UAV) technology has been widely utilized in military and civilian fields such as rescue, disaster relief, urban planning, and material transporta...Show MoreMetadata
Abstract:
Unmanned aerial vehicle (UAV) technology has been widely utilized in military and civilian fields such as rescue, disaster relief, urban planning, and material transportation. Obtaining the current position information of a UAV is of great significance in the performance of its tasks. However, positioning technology may fail in harsh environments. This article proposes the following solutions to address this problem: we eliminate interference and recover useful information by a dark channel algorithm and an attention generation adv ersarial network; visual and geomagnetic signals are fused to improve the safety of positioning technology; and a two-path convolutional network is proposed to optimize the feature extraction method and extract the continuity information of data. Moreover, open research work and challenges are also discussed.
Published in: IEEE Network ( Volume: 36, Issue: 4, July/August 2022)