Smart Cameras and Integrated Framework for Intelligent Transportation Systems

Prolonged daily periods of road traffic congestion waste time, money, and degrade both the environment and our quality of life. In Egypt, the problem is significant with severe traffic delays and high accident rates leading to devastating effects on the economic growth and challenging any progression towards sustainable development. Conventional traffic management strategies have traditionally focused on the expansion of the transportation network capacity such as building new highways and bridges. The TraffiSense project aims to provide an integrated system for traffic management in challenging road traffic conditions as customary the case is in developing countries with chaotic driving patterns, crowdedness, and lack of regulations. The proposed system will facilitate key traffic management tasks such as (a) traffic network monitoring to sense and enumerate the activities of vehicles and detect law violations; (b) transformation of raw traffic data into useful information such as congestion levels and traffic stream parameters; and (c) provision of real-time traffic information to traffic management and law enforcement authorities as well as traffic network commercial end users. To achieve its vision, TraffiSense will be exploiting and advancing state of the art algorithms in areas of visual computing, machine learning, embedded systems, and intelligent transportation systems. The development part of this phase of the project will have three key technical activities:

  •   First  ̶  development of compact computer vision algorithms for in-situ acquisition and processing of visual traffic data to produce vehicle and license plate information.
  •   Second  ̶  the construction of product-grade smart traffic sensor node device, middleware, and sensor setup interface.
  •   Third  ̶  develop advanced traveler information system that exploits information gathered by the traffic sensors and spatial dependency analysis to estimate network-wide traffic states.

The successful execution of the project will result in significant impact on the society by providing authorities with much-needed information related to traffic flow and congestion patterns to be used for managing traffic, discovering problems and planning solutions. It will also provide individual end-users with access to real-time traffic information to be used for navigation. Furthermore, the traffic information generated by TraffiSense is valuable content that can be provided to businesses and would constitute a major source of revenue for mobile operators and service providers working in traffic information publishing.











































 







Detection and segmentation of vehicles and motorcycles from videos.



The current implementation of the traffic information engine receives real-time traffic data from the sensors and utilizes this information as well as spatial dependency analysis to provide ATIS services including smart navigation and low carbon footprint journeys.




Deployment of sensor prototype at the Cairo-Alexandria desert road. The sensor detects, segments, counts, and classifies vehicles of different types (Motorcycle, Sedan, Microbus, Bus, Lorry).




The GIS-based innovative 3D visualization of traffic information to be developed by the TraffiSense project enables monitoring traffic conditions of vast areas of major roads network while plotting traffic information such as congestions and accidents blocking some roads.



 
Copyright © 2010-2014 Ubiquitous & Visual Computing Group