Head Movement Recognition Technology Research In Augmented Reality Application

Augmented Reality, which is an important branch of Virtual Reality Technology, is a transition between the real and virtual environment. Compared with the expensive hardware devices, cumbersome data and complicated computation process of Virtual Reality Technology, Augmented Reality Technology is more practical and balances both the user experience and the modeling workload. Head Movement Recognition Technology is mainly used in the multimodal human-machine interface and the remote interactive environment. The technology has broad application prospects.In this paper, we study the Head Movement Recognition Technology based on augmented reality, and realize real-time detection and identification of human head state and action using a simple video acquisition device for dynamic video. In order to improve the precision and stability of the system and to guarantee the real-time performance of the system, we do a series of optimization operation to the inputted image, including the transformation of color spaces, illumination problems and image smoothing. This paper adopted a face detection algorithm based on skin color model to locate the face, and two improvements were made on the model. On the basis of the original algorithm, we first nonlinearly transform the color space, and then classify and process the different areas according to their lighting conditions. The main purpose of the improvement is to weaken the influence of the complicated illumination conditions on the face skin color without affecting the efficiency of the system at the same time. After getting the face region, the system adopts a kind of direct and rapid method to position the eyes and mouth based on color features, to extract and mark out the facial feature points. Finally, we define a series of head movements according to the geometrical characteristics of human face location and feature points, associate these actions with some specific operations, and implement some specific functions.In this paper, a head gesture recognition system is implemented. The system can collect and initialize the user’s head and facial information through the video acquisition device, define a series of head movements, and associate the movements with some corresponding operations according to the set of users, in order to easy users operation and to enhance user experience.