Implementation of Crosswalk Lights Recognition System for the Blind's Safety
Citations

WEB OF SCIENCE

0
Citations

SCOPUS

1

초록

Recently, disabled people are exposed to various risk of traffic accidents, but, at the time of this writing, the installation rate of sound signals for the visually impaired is very low, at 57% in South Korea. In this paper, we focused on several recent services that help the visually impaired to walk safely and try to help the visually impaired through voice guidance. In this paper, we proposed crosswalk lights recognition system for the visually impaired providing real-time lights state by voice so that safe walking environment of the visually impaired can be guaranteed. We implemented crosswalk lights recognition system prototype using IoT parts like Raspberry Pi. Unlike the conventional sound guidance, we tried to provide voice guidance through crosswalk lights image recognition. The performance evaluation achieved a recognition success rate of 92.7% by day and 67.3% by night. © 2019 IEEE.

키워드

Crosswalk Light DetectionObject DetectionRaspberry PiVisually ImpairedVoice GuidanceCrosswalksImage recognitionObject detectionDisabled peopleExposed toLight detectionRecognition systemsSound signalSouth KoreaVisually impairedWalking environmentsSpeech recognition
제목
Implementation of Crosswalk Lights Recognition System for the Blind's Safety
저자
Park, HuijinWon, HeesooOu, SoobinLee, Jongwoo
DOI
10.1109/ECICE47484.2019.8942696
발행일
2019-10
유형
Conference Paper
저널명
2019 IEEE Eurasia Conference on IOT, Communication and Engineering, ECICE 2019
페이지
183 ~ 186