Vision-Based Automatic Collection of Nodes of In/Off Block and Docking/Undocking in Aircraft Turnaround

The automatic collection of key milestone nodes in the process of aircraft turnaround plays an important role in the development needs of airport collaborative decision-making. This article exploits a computer vision-based framework to automatically recognize activities of the flight in/off-block and docking/undocking and record corresponding key milestone nodes. The proposed framework, which seamlessly integrates state-of-the-art algorithms and techniques in the field of computer vision, comprises two modules for the preprocessing and collection of key milestones. The preprocessing module extracts the spatiotemporal information of the executor of key milestone nodes from the complex background of the airport ground. In the second module, aiming at two categories of key milestone nodes, namely, single-target-based nodes represented by in-block and off-block and interaction-of-two-targets-based nodes represented by docking and undocking stairs, two methods for the collection of key milestone are designed, respectively. Two datasets are constructed for the training, testing, and evaluation of the proposed framework. Results of field experiments demonstrate how the proposed framework can contribute to the automatic collection of these key milestone nodes by replacing the manual recording method routinely used today.