Automatic Real-Time River Traffic Monitoring Based on Artificial Vision Techniques

Automatic Real-Time River Traffic Monitoring Based on Artificial Vision Techniques

Luca Iocchi (Sapienza University of Rome, Italy), Luca Novelli (Archimedes Logica, Italy), Luigi Tombolini (ECOTEMA Co., Ltd, Italy) and Michele Vianello (Vice Major, Municipality of Venice, Italy)
Copyright: © 2012 |Pages: 14
DOI: 10.4018/978-1-4666-1586-1.ch008
OnDemand PDF Download:
$30.00
List Price: $37.50

Abstract

Artificial vision techniques derived from computer vision and autonomous robotic systems have been successfully employed for river traffic monitoring and management. For this purpose, ARGOS and HYDRA systems have been developed by Achimedes Logica in collaboration with Sapienza University of Rome under the EU initiatives URBAN and MOBILIS for the monitoring of the boat traffic in Venice on the Gran Canal and the harbour area. These advanced systems provide an efficient automatic traffic monitoring to guarantee navigation safety and regular flow while producing and distributing information about the traffic. The systems are based on the processing of digital images that are gathered by survey cell stations distributed throughout the supervised area providing a visual platform on which the system displays recent and live traffic conditions in a synthetic way similar to radar view. ARGOS and HYDRA systems are programmed to automatically recognize and notice situations of great interest in whatever sea or land-targeted security applications including environmental, perimeter, and security control. This article describes the wide spectrum of applications of these two systems, that is, monitoring traffic and automatically tracking position, speed and direction of all vehicles.
Chapter Preview
Top

2. Overview Of The Argos And Hydra Systems

The ARGOS and HYDRA systems have been developed and are currently under operation for 24/7 all-weather day-night operation.

The ARGOS system is controlling a waterway of about 6 km length, 80 to 150 meters width, through an array of observation posts (Survey Cells). Each survey cell is composed of 3 IR/VIS optical sensors: one center wide-angle (90 degrees), orthogonal to the navigation axis, and two side deep-field cameras (60 degrees). Detected images are rectified and stitched together so as to generate a composite plain view, similar to a radar image. The resulting overall view field along the waterway could stretch over 250-300 meters end-to-end.

The HYDRA system instead covers a much larger area, since Giudecca Canale is about 300 meters wide. Moreover, buildings are further away from the Canal and boats thus have a much smaller appearance in the image space. For the HYDRA system we thus decided to configure the Survey Cell with four high resolution color cameras with 45 to 60 degrees FOV and four 320x240 thermo cameras placed in the same direction of the color ones.

Every survey cell has also a high-resolution Pan-Tilt-Zoom camera which is used to zoom and follow a particular boat. This can be done either in an automatic way by the system or through the use of a joystick.

Figure 1.

Examples of survey cells and their views

Some examples of survey cells, installed just below the roof of several buildings leaning over the Grand Canal or on some ad-hoc installations, are shown in Figure 1. The figure also shows the views from one of the Survey Cell in ARGOS and HYDRA. Notice that the HYDRA view contains both the color image and the thermo image (where temperature is represented in gray scale).

Complete Chapter List

Search this Book:
Reset