A Proposal of OCR-based User Positioning Method in Indoor Navigation System Using Unity and Smartphone (INSUS)

An indoor navigation system helps users seamlessly navigate in complex buildings. Previously, we implemented the Indoor Navigation System using Unity and Smartphones (INSUS), where we employed QR code for accurate user positioning. However, a lot of QR code sheets need to be placed in the designated area to improve accuracy. To solve this problem, text information that is recognized by an Optical Character Recognition (OCR) software, such as room numbers, resident names, and floor levels in buildings can be used to recognize the current user position. In this paper, we propose an OCR-based user positioning method in INSUS. The text information is compared with the reference data stored in the database. To verify the performance of the proposal, we conducted a series of experiments involving various OCR models, view angles, and brightness levels. The results show that PaddleOCR achieves the average accuracy of 90.31% for various brightness levels and of 95% for different viewing angles.