Fast localization method for UAV images in GNSS-denied environments utilizing image sequence relationships

Unmanned aerial vehicle (UAV) image localization can be used in global navigation satellite system (GNSS)-denied environments for UAV self-navigation. This study proposes a fast localization method for UAV images in GNSS-denied environments based on image sequence relationships. First, the LightGlue network extracts features from adjacent UAV images and combines these features with the sequence relationship of the UAV images to perform feature matching within the overlapping range. After eliminating errors, the affine transformation matrix is calculated to realize adjacent image positioning of the UAV. Experiments conducted on multiple image datasets show that the proposed method successfully completes the rapid positioning of UAV images in GNSS-denied environments, with positioning errors less than 0.5 m, indicating potential for practical applications