Development of Walking Assistant System for Visually Impaired People Using Geomagnetic Fingerprint and Braille Block Map
Development of Walking Assistant System for Visually Impaired People Using Geomagnetic Fingerprint and Braille Block Map
复制标题
利用地磁指纹和盲文块图开发视障人士步行辅助系统
DOI:
10.1109/iccae55086.2022.9762426
复制
发表时间:
2022
期刊:
影响因子:
--
通讯作者:
Hisashi Miyamaru
中科院分区:
文献类型:
--
作者:
Koichi Kuzume;Ryosuke Nitta;Tomonari Masuzaki;Haruko Masuda;Hisashi Miyamaru
GPS-based navigation systems have been developed to support the walking of visually impaired people. However, their use in station premises and malls presented difficulties due to radio wave interference. In such places, an indoor navigation system (IDNS) was indispensable. Recently, an IDNS using geomagnetism that did not require the installation of new infrastructure attracted our attention. It incorporated technology to detect self-position by association with a geomagnetic fingerprint (GMFP) database, which included the geomagnetic vector of the current position, stored in advance on the cloud. However, to realize navigation over an entire floor, a detailed geomagnetic map of the floor was needed, taking a lot of time and labor to measure the geomagnetic field. In this research, we proposed a novel method that limited the area of the GMFP to places above braille blocks and linked the geometric maps to braille block maps. The main features of our system were as follows: (1) Reduction of labor and time required to make the GMFP; (2) Realization of a real-time navigation system with little error in estimated position; (3) A high-resolution geomagnetic map database could be easily made by walking with a nearly fixed stride length without any specific equipment. To confirm the principal operation of our system, we made a prototype system. We then evaluated the performance of our system under various experimental conditions. Finally, we demonstrated a simple navigation system. As a result, we were able to confirm that our system could estimate self-position in real-time with little error and provide simple navigation for visually impaired people.