In this paper, a graph-based method for generating structural descriptions of stereo image pairs and their matching to recover 3-D form of objects in space, is discussed. The method comprises the following steps: firstly, straight line segments are extracted in both images and the relations among them are investigated. The relational subgraphs are constructed then, and the correspondence graph is created. Using "stable marriage" searching algorithm the features in one image corresponding to features in other image are found. The method was tested on a set of aerial stereo images of urban scenes.
Keywords: structural stereo matching, computer vision, graphs