Development of Pupil Detection in Eye Movement with Raspberry Pi

Abstract

Human Computer Interaction (HCI) is an evolving technology. Eye movement is one of the method of HCI that plays an important role for disease identification and retail advertising effectiveness. Current system has types of movements that can be classified to fixation, rotation and drift, which based on electronic, mechanical or optical or video based movements. Moreover, the eye is authentic and moves faster than input media. In fact, no training is required for normal users.  However, the equipment is still expensive. Therefore, a cost effective prototype of pupil detection for eye movement is proposed. This paper introduced a development of pupil detection using fixation length of distance with video based eye movements using Raspberry Pi and web camera. The implementation of software and hardware installations are using Haar Cascade Technique. The system detects the movement of eye and capture the image of pupil using a webcam. Later, the image of pupil is compared at the matching process using String Array Concatenation on Raspberry Pi. The impact of this study is to provide a lower cost device for pupil detection in eye movement for human computer interaction system.