Image Re-Segmentation - a New Approach Applied to Urban Imagery

This article presents a new approach for image segmentation applied to urban imagery. The proposed method is called re-segmentation because it uses a previous over-segmented image as input to generate a new set of objects more adequate to the application of interest. For urban objects such as roofs, building and roads, the algorithm tries to generate rectangular objects by merging and cutting operations in a weighted Region Adjacency Graph. Objects whose union generate larger regular objects are merged or otherwise cut. In order to verify the potential of the method, two experimental results using Quickbird images are presented.