Abstract:
Since traditional obstacle detection and navigation system cannot fulfill the needs of blind, multiple electronic travel aids are introduced during past decades. However, most of them are not practical for utilizing in real life due to their size, weight, and complexity. This research proposes a framework of indoor navigation for visually impaired people by introducing 3 main parts, which are indoor positioning, obstacle detection, and way finding respectively. For positioning part, a model for differentiates location of user based on smartphone and Wi-Fi is introduced. For obstacle detection part, a prototype of obstacle detection device based on ultrasonic sensor is introduced. For way finding, an application for navigation and vibrating wristband are introduced. The main objective of the framework is to reduce cognitive load during traveling in order to promote independent traveling for blind. This research mainly focuses on positioning part by proposing a localization method based on Wi-Fi. Four access points are suitable for room level classification with the area size of 10 x 12 meter. Applying Kalman filter to raw RSSI value before generating the model increase classification performance by 6% on the average
Thammasat University. Thammasat University Library