Real-Time Image Processing Based Obstacle Avoidance System and Navigation System in Uncertain Environment for Autonomous Wheelchair Application

Lim, Jay Ming (2017) Real-Time Image Processing Based Obstacle Avoidance System and Navigation System in Uncertain Environment for Autonomous Wheelchair Application. Final Year Project (Bachelor), Tunku Abdul Rahman University College.

Text Lim Jay Ming Full text.pdf Restricted to Registered users only Download (4MB)

Abstract

The purpose of this project is to design a navigation system and an obstacle avoidance system for an autonomous wheelchair application. The obstacle detection system is done merely using the camera sensor with the knowledge of image processing. The techniques used are including Canny Edge Detection, Erosion and so on where algorithms are developed to identify and detect the obstacle. The navigation system is performed with the use of compass sensors where the compass calibration, compass reading and compass value comparison are done to ensure the autonomous wheelchair move in a correct direction towards the targeted destination. By taking inputs from both the obstacle detection system and navigation system, the motion control unit acts depending on the conditions and commands given. A pair of DC brushed motors and a pair of Smart Drive 40 motor drivers are used to perform the motion control system. The integration of these three system is meant to allow the autonomous wheelchair to reach the targeted destination safely and also to make the autonomous wheelchair system complete. Experiments on external factor(s) and internal factor(s) which may cause the performance of the autonomous wheelchair have also been conducted in this project such as the lighting conditions, the blurring coefficient, the erosion kernel matrix size and the threshold offset. The results of the experiments are analysed in order to obtain the optimum value of each parameters (internal) where the external factor(s) affect the least to the performance of the autonomous wheelchair.