Real Time Gesture Detection Using Convolutional Neural Network

Gestures make touch less connectivity and systems easier to use. In many applications, such as Augmented Reality (AR) and Virtual Reality (VR), computer vision is crucial. Systems created for people with physical disabilities and sign language detection rely heavily on gesture detection. In this study, Convolutional Neural Networks (CNN) are used to demonstrate a real-time system that enables efficient feature extraction for gesture detection. A CNN is trained for certain gestures that are essential in Human Machine Interaction (HMI), primarily for persons who struggle to communicate vocally. The performance of thresholding and depth perception, in addition to gesture detection, enables reliable results to be produced even in noisy backgrounds. Effective background elimination and contours aid in properly mapping hand gestures in a diverse range of backgrounds.