Abstract:
Detection of obstacles during navigation of a mobile robot is considered difficult due to varying
nature in terms of size, shape and location of the obstacle as well as ambient light that interferes
with infra-red (IR) signals of the robot. In this paper, we propose a novel low-cost method that
successfully guides a robot along a path using image processing and IR sensor circuits. A high
continuous IR signal is converted into a low continuous IR signal by means of a demodulation
circuit that enables a peripheral interface controller to receive this low continuous IR signal and
take relevant decisions based on the signal. The images taken from a web camera are
preprocessed to remove noise and detect edges. Subsequently, an image processing routine
effectively calculates the angle to be rotated of the front wheels using a scan line algorithm. A
minimum mean distance error of 2.45 was observed in tracking the path at a signal-to-nose ratio
of 26.50. The accuracy of speech recognition was 92% for two voice training sessions.
Description:
Kasun K. Jinasena
kasun@sjp.ac.lk
Faculty of Applied Science
University of Sri Jayewardenepura
Gangodawila, Nugegoda, Sri Lanka
Ravinda G.N. Meegama rgn@dscs.sjp.ac.lk
Department of Statistics & Computer Science
Faculty of Applied Science
University of Sri Jayewardenepura
Gangodawila, Nugegoda, Sri Lanka