A Real-Time Visual Card Reader for Mobile Devices
Abstract
With the increasing number of shops and service providers offering loyalty cards and attractive benefits for card holders, the amount of different cards a typical customer owns has become a problem. In most cases, a card needs to be scanned in the particular shop to gain discounts meaning the customers have to always carry a large number of cards. In this paper, we present an approach to automatically detect the card provider, card type and customer number, which are then stored within a mobile application, allowing customers to carry all their cards with them on their smartphone. We propose ACC* as a new evaluation metric and evaluate the performance of this approach on a large real-world dataset of loyalty cards.
Cite
Text
Stehr et al. "A Real-Time Visual Card Reader for Mobile Devices." IEEE/CVF Winter Conference on Applications of Computer Vision, 2016. doi:10.1109/WACV.2016.7477668Markdown
[Stehr et al. "A Real-Time Visual Card Reader for Mobile Devices." IEEE/CVF Winter Conference on Applications of Computer Vision, 2016.](https://mlanthology.org/wacv/2016/stehr2016wacv-real/) doi:10.1109/WACV.2016.7477668BibTeX
@inproceedings{stehr2016wacv-real,
title = {{A Real-Time Visual Card Reader for Mobile Devices}},
author = {Stehr, Lukas and Meusel, Robert and Kopf, Stephan},
booktitle = {IEEE/CVF Winter Conference on Applications of Computer Vision},
year = {2016},
pages = {1-7},
doi = {10.1109/WACV.2016.7477668},
url = {https://mlanthology.org/wacv/2016/stehr2016wacv-real/}
}