You are here

TRAVEL - Traffic Road Analysis by Visual Event Labelling

We perform automatic analysis of traffic sequences from static or moving cameras, aiming at the detection, classification and tracking of vehicles on the road. TeV played an important role is several traffic video analysis projects:

  • VIABILITA - VIdeo ABILIty for Traffic Analysis (Queue analysis formation in tunnels) 
  • SCOCA - A flexible system for counting, classifying and tracking automatically vehicles in road intersections - Partially sponsored by Comune di Trento.
  • TAS - Traffic Analysis System for risk evaluation (inside Progetto Pilota) -  partially sponsored by Ministero delle Infrastrutture e dei Trasporti in collaboration with Comune di Trento.
  • DIPLODOC - DIstributed Processing of LOcal Data for On-line Car services (FU-PAT Project) - Partially sponsored by Provincia Autonoma di Trento, in partnership with Centro Ricerche FIAT, Trento (CRF) and University of Trento, Dipartimento di Informatica e telecomunicazioni. Traffic analysis from on-board stereo camera.
  • SCOCA Mobile - Mobile SCOCA system to acquire sessions of traffica data in specific locations - Partially sponsored by Comune di Trento.
  • RIVA Tunnel - Partially sponsored by Provincia Autonoma di Trento.
  • SNODI - Travel time estimations and statistics in roundabouts - Partially sponsored by Provincia Autonoma di Trento.
Goals: 
There are two main applicative goals of visual traffic analysis we performed: (1) The collection of traffic data (volume, categorization, speed, paths) for statistical purposes and provide a base for traffic engineering decisions or to inform timely road users. (2) The identification and automatic extraction of visual characteristics of road traffic that are related to dangerous situations to identify risk before accidents appeaning.