| Summary: | 碩士 === 國立高雄第一科技大學 === 資訊管理研究所 === 100 === In this study, we have developed an augmented reality based walking system which combined with video image processing and the superposition of the 3D model of the campus tour video. User can walk in the room and simulate a walking guide to the situation on campus on the screen with the use of the proposed system. The operating process is via gesture recognition and full of interest. The proposed system consists of four part: background registration, body location targeting, footsteps detection and augmented reality technology. Background registration is achieved by using the foreground and background subtraction and combination of the adaptive threshold of body position separated from the background. Body location can calculate the symmetry axis of captured images and track body positions. Footsteps detection calculates walking steps of users by detecting the occupied ratio of hot regions which are predefined by the system.
Experimental results show that, in the guiding process, the correct rate of pace computation is up to 95%, indicating the high stability of the system and feasibility. The situation of interaction with a 3D model and superimposed with the tour video are then rendered on the computer screen. In addition, the character position in the guide is marked on the map to achieve the complete effect of augmented reality.
|
|---|
