| Computer vision technology is becoming one of the current hot research fields along with the fast development of computer technology. It can help people solve the actual questions quickly and accurately. So, the research and application of computer vision field is of great significance to the social development and people’s living standards.Augmented Reality of the mobile side has a widely application scenarios in many areas. But it now has less application on the tourism area; most augmented reality application on the tourism is mainly applied on the sign hint, navigation, and similar functions. However, one of the main significance of augmented reality technology is letting a person can have more strong immersive to reality. And many scenic spots of tourist site are not open to the public. So we try to virtual ourselves, fused with the realistic scene, make it seems like we’re really in those inaccessible spots, took photos, enhance the memory of the scenic spot.Tracking registration algorithm is the key technology to realize augmented reality system, and most of the research of augmented reality is the research of tracking registration algorithm. But in the mobile side, such as Android smart phone, the calculation processing and storage capacity are poorer than the PC side, so, if still wants to maintain good tracking registration effect and quality user experience under this circumstances, it need to guarantee accuracy at first and then try to improve the operating rate of the system and reduce the calculation of the system. Taking this as the target, this article is focus on the analyze and research of the augmented reality technology of mobile side deeply, and propose the improvement measures of corresponding tracking registration scheme and the algorithm, finally solve the related problems of augmented reality on mobile side.This article research natural feature point recognition algorithm of the nonlinear scale space method deeply by analyzing and discussing the tracking registration technique that based on natural feature point applied in the mobile augmented reality. Furthermore, this article propose related algorithm optimization and improvement measures to adapt to the calculate capacity and speed and to improve its robustness and real-time performance. |
