Development Of Obstacle Detection System For Collision Avoidance

The traditional walking stick is a common tool that visually impaired use to detect an obstacle in front of knee level of them. Nevertheless, the main problem with the use of the traditional walking stick is its inability for a user to detect obstacle above the ground level and beyond the stick dete...

Full description

Bibliographic Details
Main Author: Ng, Lai Boon
Format: Monograph
Language:English
Published: Universiti Sains Malaysia 2017
Subjects:
Online Access:http://eprints.usm.my/53021/1/Development%20Of%20Obstacle%20Detection%20System%20For%20Collision%20Avoidance_Ng%20Lai%20Boon_E3_2017.pdf
Description
Summary:The traditional walking stick is a common tool that visually impaired use to detect an obstacle in front of knee level of them. Nevertheless, the main problem with the use of the traditional walking stick is its inability for a user to detect obstacle above the ground level and beyond the stick detect range. There have many electronic travel aids (ETAs) are available for assisting mobility difficulties of the visually impaired. However, some of the ETAs exist limitation in obstacle detection for collision avoidance. To reduce the limitations of existing ETAs and mobility difficulties of visually impaired, an obstacle detection system needs to be developed. This project presents the development of a walking stick prototype with a sensor fusion based obstacle detection system. The placement and orientation of the sensor on the walking stick are successfully analysed in terms of detecting angle and range. Moreover, a tilt angle control system is implemented on the prototype depend on the height of user to ensure the sensors work effectively in the detection range. After several tests have been conducted, the tilt angle control system is able to retain the angle of the ultrasonic sensors in around the optimal angle which is 10 degrees. Lastly, a voice alert Android application is developed for assisting the visually impaired to predict the position of obstacle. There has more than 85% correct decision ratio was obtained for the high, middle, low and hanging obstacle.