Gesture Controlled Virtual Mouse and Keyboard Using OpenCV

A gesture-controlled computer makes it easier to connect with computers by utilizing hand gestures, voice commands, and eye movements. In this gesture approach computers seldom ever require direct contact, static and dynamic hand gestures, voice assistance, and eye motions can be used to virtually control all operations. In the existing systems it uses different parameters for gesture control such as voice, eye and hand. In this research paper, we are implementing a hybrid model used to perform gesture controlled by using voice, hand and eyes. We are implementing a Convolutional Neural Network (CNN) model by using Media Pipe running on top of pybind 11. The webcam, which is the only piece of hardware mandatorily required in the proposed algorithm, is used to record images, recognize hand gestures, and eye movements, and receive voice instructions using the Pyttsx3 module. It has been observed that the proposed algorithm provides more efficient results. it provides different form of gestures under hybrid model with only webcam as the input resource, it can be utilized for smart security and surveillance system.