A Low-cost Vision-based Human-computer Interface for People with Severe Disabilities

Without any doubt, mouse and keyboard are the most prevalent and important human-computer interfaces. Unfortunately, people with physical disabilities cannot enjoy benefits provided by computers as able-bodies people on equal term since conventional mouse and keyboard are designed with the able-bodies in mind. Therefore, how to lower or even tear down the barriers between computers and people with disabilities is a demanding and challenging task. Recently, advances in hardware and software have led to assistive technology systems of every variety [132]. These various systems may allow people with disabilities to use their limited voluntary motions for communications, manipulating computers, and controlling home applications (e.g. TV and air conditioner, etc.) Each system has its own considerations, applicability and limitations. Traditional interfaces fall into one of three categories: switches, joystick controllers, and pneumatic devices. Recently, advances in electronic technologies and high-speed microprocessors have led to a new generation of interfaces for the disabled. Some of these interfaces usually provide few available controlled movements and require substantial motor coordination. Furthermore, they tend to be highly specialized and are generally cost prohibitive. Since speech is a natural communication means for human being, the voice-operation feature is the idea control method for the large number of disabled people who cannot use conventional computer interfaces. Currently, several consumer speech recognition products (e.g. IBM's Via Voice and L and H Dragon NaturallySpeaking, etc) are available. They can allow users to get computer tasks (e.g. send e-mail, dictation, and surf the Web) completed quickly and efficiently. With respect to text entry, these packages may be to some extent appealing and effective. Yet full access to computers is still not completely achieved for users with disabilities since text entry does not mean everything on computers. To have full access to computers, the user needs not only be able to enter text but also access to some other applications software, the computer desktop, etc. These speech recognition products may have problems of distinguishing ABSTRACT