Smartphone-Based Accurate Self-localization by Referring to Road Sign Inventory Databases
Accurate self-localization is crucial for an increasing number of location-based services (LBS) in smartphone-based applications (i.e., APPs). This paper proposed a novel smartphone-based self-localization method by referring to road sign inventory databases. A user is expected to use smartphone to shoot any road sign nearby to collect both image and GPS/WIFI/mobile network data. The authors proposed a multi-scale localization strategy for accurate position computation. First, coarse localization is achieved from GPS or WIFI or mobile network data such that a set of road signs within the ranges are selected as the candidates. Second, the authors developed a sign recognition method to identify the matched sign within the candidates for sign-level localization. For fast recognition on smartphone, the authors proposed using ORB and SURR local feature descriptors to compute sign holistic features and Hybrid KNN for sign recognition. Finally, the authors proposed metric localization to further refine the localization results. The proposed method has been implemented into an APP on Android systems. The proposed method has been tested in two typical scenarios by using different types of smartphones. The results demonstrate that the proposed method can achieve sub-meter localization accuracy by referring to the developed sign inventory databases with conventional smartphones. The proposed method also suggests a promising way to collect updated crowd-sourcing road asset data to enhance road asset re-inventory and management.
-
Supplemental Notes:
- This paper was sponsored by TRB committee ABJ60 Standing Committee on Geographic Information Science and Applications.
-
Corporate Authors:
500 Fifth Street, NW
Washington, DC United States 20001 -
Authors:
- Li, Na
- Hu, Yuezhi
- Hu, Zhaozheng
-
Conference:
- Transportation Research Board 96th Annual Meeting
- Location: Washington DC, United States
- Date: 2017-1-8 to 2017-1-12
- Date: 2017
Language
- English
Media Info
- Media Type: Digital/other
- Features: Figures; References;
- Pagination: 14p
- Monograph Title: TRB 96th Annual Meeting Compendium of Papers
Subject/Index Terms
- TRT Terms: Accuracy; Inventory; Location based services; Pattern recognition systems; Position fixing; Smartphones; Traffic signs
- Subject Areas: Data and Information Technology; Highways;
Filing Info
- Accession Number: 01627805
- Record Type: Publication
- Report/Paper Numbers: 17-03143
- Files: TRIS, TRB, ATRI
- Created Date: Feb 27 2017 5:12PM