Realization Of Image Recognition System Based On Argumented Reality

Augmented Reality (Argument Reality, AR), also known as mixed reality, a technology proposed in the nineties, is an important branch of virtual reality technology. It calculates and tracks the position and orientation of the camera in real time, and correspond the virtual information and the real world together on display, which allows users to have the same feeling. Thus, it will effectively enhance the users’ understanding of the reality scene, interactive and enjoyment.In this paper, we discuss and analyze augmented reality system for image recognition problems, and study the local sub-invariant feature extraction process, research effective identification strategies combining existing matching algorithm and eliminate the wrong matching points, enabling the system to achieve the required timeliness and accuracy. Due to the default template is not sufficient for the case of outdoor scenes, a target segmentation method is proposed in this paper. We broadly determine the target with the GPS technology, and distinguish the goals which require enhanced display through effective segmentation technique, and the goal information getting from the network will display in the right position after rendering.Experimental results show that the proposed method can achieve the efficiency and accuracy which the augmented reality system requires. Combining with KD-tree algorithm, we achieve the system in mobile intelligent terminal, which can meet the needs of expanding information to some extent, and it has a wider range of application than the traditional one.