Real-Time Processing of High-Resolution Video and 3D Model-Based Tracking for Remote Towers

  • PDF / 1,377,141 Bytes
  • 10 Pages / 595.276 x 790.866 pts Page_size
  • 72 Downloads / 145 Views

DOWNLOAD

REPORT


ORIGINAL RESEARCH

Real‑Time Processing of High‑Resolution Video and 3D Model‑Based Tracking for Remote Towers Oliver J. D. Barrowclough1   · Sverre Briseid1 · Georg Muntingh1 · Torbjørn Viksand1 Received: 5 May 2020 / Accepted: 14 August 2020 © The Author(s) 2020

Abstract During the past decade, a new approach to providing air traffic services to airports from a remote location has been established, known as remote or digital tower. High quality video data is a core component in remote tower operations as it inherently contains a huge amount of information on which a controller can base decisions. The total resolution of a typical remote tower setup often exceeds 25 million RGB pixels and is captured at 30 frames per second or more. It is thus a challenge to efficiently process all the data in such a way as to provide relevant real-time enhancements to the controller. In this paper we describe the development of number of improvements and discuss how they can be implemented efficiently on a single workstation by decoupling processes, implementing attention mechanisms and utilizing hardware for parallel computing. Keywords  Remote tower · Object tracking · Machine learning · Video processing · 3d Modelling

Introduction The remote tower concept, proposed in the early part of this century, aims to provide air traffic services (ATS) to airports from a remote location by streaming sensor data over a network [21, 29]. High-resolution video that replicates the out-the-window view of a conventional tower is a core feature of the remote tower concept and is supplemented by a pan-tilt-zoom (PTZ) camera that mimics the function of binoculars in a conventional tower [8]. Other components in remote tower typically include voice communication, meteorological data, sound reproduction and radar data, where available. The motivation behind the concept is both to cut costs associated with tower maintenance and personnel, and to improve safety. Today there is a focus on validating the multiple remote tower concept, in which an air traffic

* Oliver J. D. Barrowclough [email protected] Sverre Briseid [email protected] Georg Muntingh [email protected] Torbjørn Viksand [email protected] 1



controller (ATCO) is responsible for more than one airport simultaneously [16, 22]. This movement of the concept towards more complex scenarios demands new technologies that support an ATCO’s situational awareness in order to ensure that their cognitive capacities are not exceeded. Such technologies may include providing visual enhancements or even fully automating parts of the ATCO’s responsibilities. This requires processing of large amounts of data, including high-resolution video data, in real-time. In this work we consider up to 14 statically positioned cameras, each with resolution 1080x1920 aligned in portrait mode and stacked horizontally, for a total resolution of 15120x1920. Despite the ever-increasing capacity of modern computers, their computational power is still insufficient to allow for naive imp