Abstarct: In this research, a Human-Machine Interface has been designed which can turn people with severe disabilities able to communicate with the other people. This HMI can track and analyze the user’s eye motions; by this, the users can access the HMI software different buttons and by selecting the buttons, they can express their needs. The user’s eye motions can be likened to the computer mouse motions. In the mentioned HMI, at the first step, the user’s eyes are detected by using a HAAR cascade classifier algorithm. As the eyes pupils are circles, at the second step, a Hough transform has been employed to detect the pupils. By performing several tests, three threshold values have been specified in three directions (Upward, right and left); The Hough circle center position is analyzed simultaneously and by using the mentioned thresholds and the Hough circle center position, the user’s looking direction is detected. The user’s optional blxinking can be likened to the computer mouse left click and the user can select different buttons with optional blxinking; in this regard, if the Hough circle cannot be detected in twenty continuous frxames, the HMI interprets this action as the optional blxinking. The HMI has been tested by thirty different users; the tests results show that the HMI is working properly.