A new method to combine detection and tracking algorithms for fast and accurate human localization in UAV-based SAR operations

This paper presents the study and the evaluation of GPS/GNSS techniques combined with advanced image processing algorithms for the precise detection, positioning and tracking of distressed humans. In particular, the issue of human detection on both terrestrial and marine environments, as the human silhouette in a marine environment may differ substantially from a land one, is addressed. A robust approach, including an adaptive distressed human detection algorithm running every N input image frames combined with a much faster human tracking algorithm, is proposed. Real time or near-real-time distressed human detection rates, under several illumination and background conditions, can be achieved using a single, low cost day/night NIR camera. It is mounted onboard a fully autonomous UAV for Search and Rescue (SAR) missions. Moreover, the collection of a novel dataset, suitable for training the computer vision algorithms is also presented. Details about both hardware and software configuration as well as the assessment of the proposed approach performance are discussed. Last, a comparison of the proposed approach to other human detection methods used in the literature is presented.