Scholarlink

The Smart Sunglasses for Visually Impaired Individuals is an IoT-enabled assistive system designed to enhance users’ safety, mobility, and independence. The system integrates a Raspberry Pi Zero 2 W, camera, radar sensor, GPS module, and audio feedback system, all housed within a 3D-printed sunglass frame. It provides real-time object detection, live location tracking, and emergency notification features by employing the SSD MobileNet algorithm (Liu et al. in Computer vision—ECCV 2016. LNCS. Springer, Cham, pp. 21–37, 2016) trained on the MS COCO dataset (Lin et al. in Computer vision—ECCV 2014. LNCS. Springer, Cham, pp. 740–755, 2014) for obstacle identification. A Firebase real-time database (Firebase in Firebase realtime database, 2025; Firebase Team in Google: firebase realtime database: data synchronization and security, 2025) ensures secure, synchronized data storage, while a Flutter-based mobile application allows permitted guardians to monitor the user’s location and receive alerts during emergencies. The design also emphasizes power efficiency, lightweight ergonomics for user comfort, and privacy protection through encrypted data handling. System evaluation showed obstacle detection accuracy of 89.2% in daylight and 73.8% under sufficient artificial light at night, with reduced latency of 380 ms. With a focus on portability, scalability, and user-friendliness, the proposed solution demonstrates its potential to address accessibility challenges and empower visually impaired individuals with greater confidence in navigation and safety.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

IoT Based Navigation and Safety System for Visually Impaired Person

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Md Moshiur Rahman,
  • Md Shihab Hassan Naim,
  • Md. Tahzib Ul Islam

摘要

The Smart Sunglasses for Visually Impaired Individuals is an IoT-enabled assistive system designed to enhance users’ safety, mobility, and independence. The system integrates a Raspberry Pi Zero 2 W, camera, radar sensor, GPS module, and audio feedback system, all housed within a 3D-printed sunglass frame. It provides real-time object detection, live location tracking, and emergency notification features by employing the SSD MobileNet algorithm (Liu et al. in Computer vision—ECCV 2016. LNCS. Springer, Cham, pp. 21–37, 2016) trained on the MS COCO dataset (Lin et al. in Computer vision—ECCV 2014. LNCS. Springer, Cham, pp. 740–755, 2014) for obstacle identification. A Firebase real-time database (Firebase in Firebase realtime database, 2025; Firebase Team in Google: firebase realtime database: data synchronization and security, 2025) ensures secure, synchronized data storage, while a Flutter-based mobile application allows permitted guardians to monitor the user’s location and receive alerts during emergencies. The design also emphasizes power efficiency, lightweight ergonomics for user comfort, and privacy protection through encrypted data handling. System evaluation showed obstacle detection accuracy of 89.2% in daylight and 73.8% under sufficient artificial light at night, with reduced latency of 380 ms. With a focus on portability, scalability, and user-friendliness, the proposed solution demonstrates its potential to address accessibility challenges and empower visually impaired individuals with greater confidence in navigation and safety.