Many people use alternative input devices (other than standard keyboard and mouse) to access computers due to mobility impairments or other disabilities including cerebral palsy, ALS, stroke, spinal cord injury, or repetitive stress injury. The ready availability of powerful personal computers and inexpensive digital video cameras create the opportunity to develop a new type of practical computer interface: gesture recognition. This project focuses on recognition of multiple head and face gestures. The system is designed to use standard low cost digital video cameras (under $100) and standard personal computers. Face and head gestures are used to generate mouse or keyboard actions to control the computer. For example, one application to be developed is the capability to "surf the web" hands-free using head gestures to navigate web pages, including selecting and activating desired links. Gestures recognition can be used to replace or augment existing switches (such as head, foot, or hand activated switches) to interact with a computer. This has the potential to greatly improve speed of computer access. Gesture recognition is expected to fill an important gap between switches, speech recognition (which has a number of disadvantages in work and school settings) and expensive head tracking systems for computer control. Summary of Anticipated Results and Implications: The goal of this project is to develop an interface to control a computer through gestures, for example winking, nodding, or shaking your head. While it is anticipated that this type of interface would be convenient and simple for all computer users, early adopters are likely to be those who would benefit most users with mobility impairments who have difficulty using standard keyboard and mouse input devices. Anticipated results of this project include development of a low cost gesture recognition system that uses existing PC hardware and digital camera hardware to recognize head gestures. The system is likely to have early applications for people with disabilities to replace or augment switch activated systems, and to act as an alternative to a pointing device for individuals with upper limb repetitive stress disabilities. Other commercial applications include use by people with repetitive stress injury to alleviate symptoms by reducing repetitive upper limb motion through use of head gestures. Future applications may include use with handheld electronics that are expected to have an inexpensive integrated digital video camera (cell phones, personal digital assistants, etc.) to provide convenient navigation and control of complex functions since these devices have only a few small buttons.
