Pairwise representation for image database indexing

This paper develops a simple and robust representation scheme for indexing line patterns. The application vehicle used in this study involves indexing into an aerial image database using a cartographic model. The images contained in the database are of urban and semi-urban areas. The cartographic model represents a road network known to appear in a subset of the images contained within the database. There are known to be severe imaging distortions present and the data cannot be recovered by applying a simple Euclidean transform to the model. The cartographic indexing into the database is achieved using features computed from neighboring line segments extracted from the raw aerial images. These indexation features are based on directed relative angle and directed relative position of segment pairs. We illustrate both the effectiveness of our representation and the advantage that it offers in overcoming local shape ambiguities.