Eye tracking system using isophote eye center detection with blink perception

Eye is the main sensing organ in human body, so it contributes much for user interface. Eye tracking has been widely used in Human Computer Interactions. Real time face image capturing along with eye gaze tracking can provide a means of user input to the computer. Here we combine an efficient method for tracking eye and detecting blink to produce an environment by which physically disabled persons can access the computer very easily. Our aim is to minimize the hardware and setup cost by using accurate image processing techniques. The system will track the eye of a person by camera and process them to control the mouse pointer on the screen. The user can control the mouse by moving and blinking the eye. The system use isophote eye center detection method to find the iris. It gives an accuracy of 99.20%. The curves that connect the points of equal intensities are called isophote. Binary equivalent image is used to perform blink detection, in which number of black pixels is counted to determine the blinking of eye. Face detection is also done in order to eliminate unwanted regions.