Road-crossing assistance by traffic flow analysis

Adi Perry*, Nahum Kiryati

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Scopus citations


We present a system to alert visually impaired pedestrians of vehicles approaching a road-crossing without traffic control. The system is computationally efficient, requires low-cost hardware, and can be mounted on existing street infrastructure, such as sign or lighting poles. The incoming video stream, showing the approaching traffic, is transformed to a one-dimensional signal, that is forwarded to a decision module. Preliminary experimental results indicate promising probabilityof- detection and false alarm rates, while providing sufficiently early warning to the pedestrian. The planned target hardware is a solar-charged low cost Android device.

Original languageEnglish
Title of host publicationComputer Vision - ECCV 2014 Workshops, Proceedings
EditorsCarsten Rother, Lourdes Agapito, Michael M. Bronstein
PublisherSpringer Verlag
Number of pages14
ISBN (Electronic)9783319161983
StatePublished - 2015
Event13th European Conference on Computer Vision, ECCV 2014 - Zurich, Switzerland
Duration: 6 Sep 201412 Sep 2014

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Conference13th European Conference on Computer Vision, ECCV 2014


  • Approaching traffic analysis
  • Blind
  • Optic flow
  • Resource-limited system
  • Road crossing assistance
  • Visually impaired


Dive into the research topics of 'Road-crossing assistance by traffic flow analysis'. Together they form a unique fingerprint.

Cite this