日本建築学会計画系論文集
Online ISSN : 1881-8161
Print ISSN : 1340-4210
ISSN-L : 1340-4210
物的手掛かりが少ない環境下での音声経路案内の有効性
スマートフォンの音声経路案内を用いた視覚障害者の歩行に関する研究 その1
内藤 拡也, 五十嵐 雄哉, 貞清 一浩, 山田 哲弥
著者情報
ジャーナル フリー

2018 年 83 巻 751 号 p. 1695-1703

詳細
抄録
 Introduction
 We developed the turn-by-turn voice navigation system for blind people with high accuracy and using their smartphone. Many navigation tools have been proposed so that visually impaired people can walk independently in unfamiliar places, but they did not have sufficient accuracy and required specialized devices. We verified the effectiveness of our voice navigation system in the environment with few physical clues

 Method
 We conducted 2 types of test: A) walking skill test and B) walking test using the navigation, with 10 blind and low vision participants.
 A) Walking skill test: we measured their walking skill such as distance sense, direction sense, walking speed and veering of straight line walking.
 B) Walking test using the navigation: we measured their walking path, walking distance and duration time in 2 areas with few landmarks (Piloti and Lobby in Fig. 2).

 Result
 The test results are as follows:
 1) The more error they had in direction sense test, the more error they had in straight line walking test (Fig. 7). It is presumed that the relation between veering and direction sense error. The more error they had in straight line walking test, the longer distance they walked in Piloti (Fig. 11). Most of the low vision participants got to the entrance directly because they found the doormat in front of the entrance. Meanwhile, some participants who tend to veer off the route, walked meanderingly although the system tried to correct their position. This result suggests the voice navigation is hard to correct their veering.
 2) When the turning point was deviated from the route, low vision participants corrected their position, confirming visually. Fig. 13 shows the low vision participants stopped walking in a short time to look for the stairs when they turned. Meanwhile, most of the blind participants corrected their position, using cane technique, but some participants didn't because turning point was too deviated to cane reach. Therefore, we consider that localization error has to be within about 1 meter that one's cane reaches.

 Discussion
 All participants succeeded in walking in the environment with few landmarks by using the voice navigation system. However, some participants missed turning or walked meanderingly because the system didn't correct their wrong positon or direction perfectly. It is assumed that we should improve not only localization accuracy but also usability by personalization based on user's walking skill.
著者関連情報
© 2018 日本建築学会
前の記事 次の記事
feedback
Top