Automatic analysis of airport processes

Coordinator: Jose Maria Gonzalez Linares

Participants: Jose Maria Gonzalez Linares, Nicolas Guil

Funded by the company Tecnologias Digitales Audiovisuales (TEDIAL), Duration: Two years (2008-2009)

Description: In this project, a video surveillance system was developed to monitor automatically different airport processes.It was integrated with the application PROMISYS by the company AERIAM. A prototype was built to detect plane arrival and departure, passenger bus arrival and departure, people flow and crowding during the check-in. Keypoints of the prototype include:

  • Storage of video sequences of different airport processes
  • Particularized analysis of several different processes
  • Robust operation under adverse weather conditions (rain, fog, solar glare, etc)
  • Annotation of different events in a database
  • Development of a tool for supervised classification of objetcs of interest in the video sequence
  • Identification of different vehicles in the airport dock

A picture of the classification tool. Every object identified in the video stream can be classified for a later training:

Classifier tool

Detection of plane arrival. The plane is identified and the event time is annotated:

Plane Arrival

Detection of crowds in check-in queues. Simultaneous queues are monitorized to detect crowding:

Chech-in queues