Research And Application Of Interactive Augmented Reality Based On Face Detection

Augmented Reality mixes objective scenes with generated virtual scene through image processing technology and visualization techniques. Advantages of great immersively interacting and entertainment make it a hot spots for research institutions and IT giants,and some of them release several AR development kits or applications on mobile platforms. However, earlier AR technology lacks of interacting methods, and the most flexible interacting media are hands and face, and image process library supports gesture recognition and face detection.In the paper QCAR and OpenCV are taken as the development kits, after simulation environment constructed on iOS platform,the processes of interacting and scene registering are simulated with the aid of OpenCV.And some key points and the implementation process are presented in detail.In the simulation, strong cascade classifiers based on Haar features are used for fast face detection,which is of great scalability.The method support face detection through training special classifier,and it can be expanded to support hand detection and gesture recognition.In4th chapter,a method is proposed for increasing the speed of face detection on mobile platforms.Then the shortages,factors affecting face detection are analyzed and the follow-up key points of research are introduced in detail.As the speed of face detection directly affects the fluency of experiencing AR applications,the research focuses on improving face detection,and from the simulation data we make a conclusion that the proposed method can increase the speed of face detection in AR application.This paper explores to provide a general,convenient and low-cost interacting method,which has a strong potential value for AR application on mobile platforms.