Real Time Distance and Mobility Detection of Obstacles Using a Smart Multisensor Framework for Visually Impaired People

Abstract

In the present era, every people wants to be self-reliant even if he is a visually impaired person. Even these people want to be self-reliant in the busy road or a crowded place. This paper offers a system that helps these people. The system includes three Ultrasonic Sonar sensors, three Passive Infrared sensors (PIR sensor), raspberry pi, SD card, vibrator and headphones. All equipment is built into a mobile cane. The PIR sensors allow the proposed system to detect movable obstacles like walking people, vehicles etc around the user in the left, the right and the front. The Ultrasonic Sonar sensors measure the distance and velocity of the movable object from the cane. Then the system notifies the user using vibration and audio message depending on the direction of the coming objects, distance and velocity. The system was experimented with real data for 7 times. The framework has given average accuracy for detecting movable object by PIR sensors is 85.9%, for distance and velocity measurement by ultrasonic sonar sensors are 91.84% and 85.04% respectively. Its accuracy is better and can be usable in real life.