Eye state detection in human faces by using circular hough transform

Nowadays eye states are used as inputs to many applications. Facial expression recognition, human-computer interaction and driver fatigue detection systems could be given as an examples to this applications. By employing these kinds of applications in academical and commercial fields, eye state detection has drown great attention in the past decade.In this thesis study, an eye state detection system has been developed which is based on detecting the circular structure that iris forms inside the eye by means of circular Hough transform. Initially, face and eye pair images are obtained by employing Viola-Jones object detection framework. Later, eye images are cropped from eye pair images that are acquired from the previous step. For extracting the iris from the sclera, image enchancement techniques like histogram equalization and thresholding were used. Sequel to this, some process has been made for putting forward the region of interest on the black and white image. With the help of canny edge detector, edges were obtained in the black and white images and these edges were used for circle detection with the help of circular Hough transform.Key Words: Eye State Detection, Circular Hough Transform