A Marker Tracking Method and a Single Hand Input Interface on a Structual Analysis for Wearable AR System


This paper proposes a new method of a marker tracking and a new input interface for a wearable computer based Augmented Reality (AR) system. Recently, as the computer technology advances, a wearable computer has been becoming widely available. Wearable computers are often deployed in Augmented Reality (AR) systems. In such systems, markers are often used for rendering virtual objects into the real space. There are various marker tracking systems already on the market. These marker systems contain enough amount of information and produce good recognition accuracy. These systems are suitable for using on an AR system. However, we observe that a marker system may have various usages if the system possesses large amount of information. Therefore, improvement in terms of the amount of information or the width of the recognizable area is needed in order to make these marker systems usable in a wearable AR system. The main feature of our visual marker is a large amount of information that it can possess. Our marker consists of 32 bits data area with error detection code. This feature enables a wearable AR system to be used in various ways. As an input interface for a wearable computer, a mouse or a keyboard are often used. However, these devices are not suitable for a input interface for a wearable AR system. In such a input interface, the hardware configuration should be minimal. It is also needed that the input methods are as intuitive as possible. In our input interface using a single hand, we determine the shape of the hand by analyzing an image. In traditinal systems, the shape of the hand is often estimated by the outline of the hand. However, it is difficult to determine the shape accurately by this method. In our implementation, we determine overlaps of the hand parts by effectively applying the edge inside the hand.