in this paper, we describe a 3-D facets construction system which relies on the simultanueous analysis of regions and contours of stereo-images in order to reconstruct 3-D indoor scenes. Firstly, a Split-And-Merge algorithm cooperating with an edge extractor is used to compute an initial region segmentation which is improved by using a rule-based system. Then, the system matches relevant regions in both images. After this, the contours detected by an edge extractor are used to replace irrelevant region boundaries after which the algorithm yields homologous subsets of contour segments. By comparing the results in each image, pairs of 2-D facets are extracted from the homologous subsets. Next, three-dimensional triangulation data are approximated by planes. Finally, planar 3-D facets are computed by an iterative algorithm to improve the initial estimation of the facets. These 3-D facets define the 3-D model of the observed scene.