Comparative Analysis of Hands-free Mouse Controlling based on Face Tracking

Abstract

Computer and mouse are two devices that inseparable from each other. Because mouse movement will control cursor movement to do any task that occurs on the computer, our research will also replace the role of the mouse in controlling cursor movement, using facial tracking by HOG and Haarcascade. Using facial movements instead of the mouse to move the cursor, users can minimize hand movements so users with impaired hands can operate the computer without a mouse. It is called hands-free. We use the HOG and Haarcascade method to determine the difference in time required by each method to control user movement. Here we experiment with 12 participants to find out the difference in time and accuracy. We use ANOVA analysis to produce a significant time difference and accuracy between those two methods. The accuracy shows that HOG has better accuracy than Haarcascade. HOG's accuracy is about 95.79%. In addition, age category analysis also affects the time generated. From this age category, it turns out that it produces a significant difference.