The Design And Implementation Of Mobile Augmented Reality System Base On Faces

In recent years, with the rapid development of mobile internet and the rapid ascension of intelligent terminal equipments' configuration and performance, mobile terminals have become the entrance of many products. The video applications have a wide range of user groups, thus there are a lot of manufacturers or individuals engaged in the development of related applications. When the mobile terminal users use handheld devices' camera preview themselves, they often want the real environment in the screen can blend some virtual scene or object to enhance the visual effect. Currently, the video applications on the market with similar function mostly use the 2d map covering or overlapping the real-time preview image to enhance visual effect, and lack of the augmented reality application which blending the real scenes and the virtual 3d objects. Therefore, this thesis focuses on design and implementation of a real-time augmented reality system of face detection and tracking based on Android platform.The content of this system mainly includes three aspects:face detection, face tracking and real-time augmented reality. In the first step, system uses Android device's camea to preview image. At the same time, system preprocesses the video frame in order to reduce the illumination influence of the image's quality. In the second step, system uses the face detection method based on Adaboost algorithm doing face detection. Once detected the the location and size of the face will be saved as the initial goal of face tracking module, and then using the method of target tracking based on Camshift algorithm doing face tracking. In the last step, system uses OpenGL ES 2.0 technology rendering 3d graphics, and overlaying in the surrounding of face according to its real-time location.3d model can translate, zoom and rotate according to the location and size of the face in real time to augment the reality.