We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
ISR: indoor shop recognition via user-friendly and efficient fingerprinting on smartphones.
- Authors
Zhao, Dong; Xu, Huaiyu; An, Jiaqi; Liu, Liang; Ma, Huadong
- Abstract
Indoor shop recognition can not only help mobile users to quickly recognize a shop to know about the information of interest without needing to enter the shop, but also assist in achieving more accurate user localization in a shopping mall. However, the existing Wi-Fi fingerprint-based approaches or image-based approaches cannot accomplish this goal well due to a huge cost of constructing large-scale fingerprint database and poor accuracy. In order to address these issues, we proposed a user-friendly and efficient fingerprinting method to collect various valuable sensory data with smartphones, which can not only reduce the randomness of fingerprints and the negative impact of pedestrians in image matching, but also be used to derive the user-to-shop distance based on the perspective projection model for assisting in determining an accurate fingerprint searching scope. We also proposed an efficient fingerprint searching and matching method to improve the recognition accuracy. We implemented a prototype system and collected fingerprint datasets in a shopping mall. Extensive experiments demonstrate that our solution achieves promising results in realistic scenarios.
- Subjects
HUMAN fingerprints; PATTERN recognition systems; SMARTPHONES; CELL phone users; WIRELESS Internet; IMAGE registration
- Publication
Machine Vision & Applications, 2017, Vol 28, Issue 7, p781
- ISSN
0932-8092
- Publication type
Article
- DOI
10.1007/s00138-017-0838-2