Real time human computer interface application based on eye gaze tracking and head detection

The localization of eye centers and the knowledge of eye movements is becoming an important search area on human-computer interactions with increasing developments in the computer vision branch. As software, obtaining the position of the small region which the user look on the screen is a simple and efficient alternative way for use of mice and touchy devices to control the cursor. In this study, firstly, Viola Jones Algorithm is used to detect eye areas through acquired webcam image on Matlab platform. Then, undistorting that is for making the eye center points more accurate, a camera calibration algorithm is included. Grey level and morphology level operations are achieved step by step. Eye centers are described using Circular Hough Transform Method. Finally, an application of kids interface is developed. Letters, objects, colors, numbers become to be learned simply and funny by this interface. The applicability of the proposed algorithm is tested and gathered results are presented.