An iterative clustering algorithm for classification of object motion direction using infrared sensor array

Abstract

Infrared sensors have been widely used in the field of robotics. This is primarily because these low cost and low power devices have a fast response rate that enhances realtime robotic systems. However, the use of these sensors in this field has been largely limited to proximity estimation and obstacle avoidance. In this paper, we attempt to extend the use of these sensors from just distance measurement to classification of direction of motion of a moving object or person in front of these sensors. A platform fitted with 3 infrared sensors is used to record distance measures at intervals of 100ms. A histogram based iterative clustering algorithm segments data into clusters, from which extracted features are fed to a classification algorithm to classify the motion direction. Experimental results validate the theory that these low cost infrared sensors can be successfully used to classify motion direction of a person in real time.