Augmented reality (AR) technology for digital composition of animation with real scenes is to bring new digital entertainment experience to the viewers. Augmented reality is a form of human-machine interaction. The key feature of the augmented reality technology is to present auxiliary information in the field of view for an individual automatically without human intervention. The effect is similar to composing computer-animated images with real scenes. To achieve the new augmented reality experience, two main problems are: How to keep track of the viewing parameters of the individual viewer? How to render a virtual image in the field of view correctly and seamlessly? To tackle the above, we are designing and implementing an enabling framework, called augmented reality computing arena for digital entertainment (ARCADE), to support the creation of augmented reality entertainment applications. Moreover, we are also developing two new video object tracking algorithms, 3D surface markers tracking and human head and face tracking. In this paper, ARCADE's system architecture and its components are described in detail and illustrated with an application example