Designing real-time vision-based augmented reality environments for mobile applications.

Augmented Reality (AR) is a variation of Virtual Reality (VR). It allows the user to see computer generated virtual objects aligned with or superimposed upon objects in the real world through the use of some kind of see-through head-mounted display. Human users of such system can interact with the virtual world and perceive information, such as character descriptions of physical objects and instructions for performing physical tasks, in form of annotations, speech instructions, images and 3D models, from input devices such as screens, microphones, or haptic devices. This thesis describes our work of building a wireless collaborative augmented reality system, which supports vision-based tracking, video-based 3D graphics, and a keyboard interface over wearable computers to interact with virtual objects. Our goal is to have all the mobile users who work in the same real world space share the predefined virtual space and interact with the virtual world in real time. Multi-player Games in AR domain and Collaborative Conferences are two typical applications. One of the most critical requirements for augmented reality applications is to recognize and locate real world objects with respect to the person's head position and orientation. In this thesis, we propose a new technique for identifying real world objects. The method utilizes a Binary Square Marker, which can identify a great number of real world objects with markers tagged on them by using computer vision techniques. The marker is a link between the real world and the virtual world. It can be detected by a CCD camera attached to a mobile device. The marker works to determine the 3D position and orientation of the object with the marker tagged on it, as well as its unique ID number. Our technique is based on the observation that given four vertices of a square marker region, the coordinates of the projection of every point on the marker can be computed as a linear combination of the coordinates of the projections of four vertices under the assumption of affine transformation