Innovation of Touchless Touchscreen Technology in Automotive User Interface

Abstract

Inside a car environment music system plays a major role for entertaining people. In this paper, a music system with touchless and vision based GUI is granted. This menu-driven UI is offered by controlling actions of the fist. Both these algorithms can be brought by the deep learning technique. Convolutional Neural Network (CNN) is used in the hand posture recognition technique to make the user interface interactive to perform the actions initiated by the user. Long-Term Recurrent Convolutional Neural Network (LRCNN) algorithm is used to ignite the touchless interface by the gestures. When a fist movement is carried out, a sequence is captured in the form of multiple image frames. So this can be accomplished using the deep learning technique. Sampled images are taken from the video sequence that is captured during the gesture recognition. Key frame extraction technique is adopted to obtain finer images from the video sequence using sparse learning. Sparse dictionary learning is used as it is individually optimized for the video sequence but, is expensive computationally.