presentationposted on 26.06.2017, 00:00 by Cody Fagley, Mark Gradecki, Wyatt Horner, Shane Firnekas
MouseVision is a project that combines optical tracking and speech recognition software to create a mouse that does not rely on the use of arms and legs. It utilizes the openCV API to find an initial point to track and creates a point vector. Then it detects changes in the point vector to create a corresponding change in the mouse position on screen. It combines this aspect with the CMU Sphinx speech recognition library to perform clicks when the end user says "left click" or "right click". The goal of this project was to make an interface that used cost-effective hardware, such as a webcam and microphone, to allow someone who is quadriplegic, or otherwise physically limited, to operate a mouse cursor. The project is considered successful, because a person without any arms at all is able to do basic computer tasks, such as play chess, using the MouseVision software.