Vision-Based UAV Localization System in Denial Environments

Abstract

Unmanned Aerial Vehicle (UAV) localization capability is critical in a Global Navigation Satellite System (GNSS) denial environment. The aim of this paper is to investigate the problem of locating the UAV itself through a purely visual approach. This task mainly refers to: matching the corresponding geo-tagged satellite images through the images acquired by the camera when the UAV does not acquire GNSS signals, where the satellite images are the bridge between the UAV images and the location information. However, the sampling points of previous cross-view datasets based on UAVs are discrete in spatial distribution and the inter-class relationships are not established. In the actual process of UAV-localization, the inter-class feature similarity of the proximity position distribution should be small due to the continuity of UAV movement in space. In view of this, this paper have reformulated an intensive dataset for UAV positioning tasks, which is named DenseUAV, aiming to solve the problems caused by spatial distance and scale transformation in practical application scenarios, so as to achieve high-precision UAV-localization in GNSS denial environment. In addition, a new continuum-type evaluation metric named SDM is designed to evaluate the accuracy of model matching by exploiting the continuum of UAVs in space. Specifically, with the ideas of siamese networks and metric learning, a transformer-based baseline was constructed to enhance the capture of spatially subtle features. Ultimately, a neighbor-search post-processing strategy was proposed to solve the problem of large distance localisation bias.

Finding Point with Image: An End-to-End Benchmark for Vision-based UAV Localization

“…Recently, Univerisity-1652 [15] introduced drone view into Cross-View Geo-localization and proposed two drone-based subtasks: drone-view target localization and drone navigation, and regarded them as image retrieval tasks. Inspired by Univeristy-1652, DenseUAV [7] achieved high precision localization of UAVs by dense sampling, which is the first time to solve the UAV localization problem by image retrieval. Table I lists the amount of training data, sampling platforms, data distribution, localization targets, and the evaluation metrics for all datasets mentioned above.…”

“…To ensure fairness, the image scales used for retrieval are consistent with FPI. Besides, In order to reduce the gap between domains, we reconstruct the training set for image retrieval in the same way as DenseUAV [7] named UL14-R. We counted RDS and inference time as shown in III. While the speed of FPI is 4 times that of the image retrieval model (FSRA-2B), it also has nearly 33 points of improvement on RDS.…”

“…DenseUAV [7] draws on the framework of Cross-View Geo-Localization to realize the visual positioning of UAVs. Thus adopted an image retrieval approach, using drone images as a query to retrieve the most similar satellite images in the entire satellite image database.…”

MTGL40-5: A Multi-Temporal Dataset for Remote Sensing Image Geo-Localization

“…With the advancements in aircraft technology, acquiring high-quality UAV and satellite RS images has become increasingly easier. Therefore, cross-view geo-localization has attracted significant attention from researchers [26,27], which focuses on the query and database images captured from different views, such as UAV ∼ satellite [19,28], ground ∼ satellite [29,30], and UAV ∼ ground [9]. In addition to the challenges posed by dynamic elements in ground images, the use of UAVs for image capture provides more freedom, resulting in diverse angles of images.…”