An AR Map Virtual–Real Fusion Method Based on Element Recognition

The application of AR to explore augmented map representation has become a research hotspot due the growing in maps and geographic information addition rising demand for automated interpretation. Taking as object, this paper focuses on tracking registration virtual–real fusion method based element recognition. It strives establish new visualization interface model. technology is applied 2D planar maps. A step-by-step identification extraction unmarked elements are designed proposed analysis characteristics elements. This combines spatial attribute point-like line-like elements, extracts color, geometric features, distribution through computer vision methods, completes automatic multi-target image recognition template contour matching, line color space area growth introduced detail. Then, 3D used realize images, algorithm proposed. experimental results an stepwise reveal that model studied effective. Through efficiency rate, it proved fast, its meets real-time requirements, accuracy high.