Road-Crossing Assistance by Traffic Flow Analysis

Abstract

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 probability-of-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.

Cite

Text

Perry and Kiryati. "Road-Crossing Assistance by Traffic Flow Analysis." European Conference on Computer Vision, 2014. doi:10.1007/978-3-319-16199-0_26

Markdown

[Perry and Kiryati. "Road-Crossing Assistance by Traffic Flow Analysis." European Conference on Computer Vision, 2014.](https://mlanthology.org/eccv/2014/perry2014eccv-road/) doi:10.1007/978-3-319-16199-0_26

BibTeX

@inproceedings{perry2014eccv-road,
  title     = {{Road-Crossing Assistance by Traffic Flow Analysis}},
  author    = {Perry, Adi and Kiryati, Nahum},
  booktitle = {European Conference on Computer Vision},
  year      = {2014},
  pages     = {361-374},
  doi       = {10.1007/978-3-319-16199-0_26},
  url       = {https://mlanthology.org/eccv/2014/perry2014eccv-road/}
}