Approximate Subgraph Isomorphism for Image Localization.

摘要: We propose a system for user-aided image localization in urban regions by exploiting the inherent graph like structure of streets, buildings and intersections. In this nodes represent buildings, intersections roads. The edges “logical links” such as two being next to each other, or building on road. generate automatically large areas using publicly available road footprint data. To localize query image, user generates similar manually identifying roads image. then run subgraph isomorphism algorithm find candidate locations evaluate our multiple sizes ranging from 2km 47km Amman,Jordan Berkeley,CA,USA. have found that many cases we reduce uncertainty query’s location much 90 percent.