Images Matching Using Voronoï Regions Propagation

Abstract

This paper presents a robust dense matching algorithm based on a geometric approach Voronoï. Feature points are matched and used to divide the images (left and right) to Voronoï regions. A left Voronoï region corresponds with its right counterpart, if the sites of the two regions constitute a true matching (seed).The two regions have the same number of points, the same shape and are highly correlated. The points of the Voronoï regions which satisfy all these criteria serve as new seeds for the next iteration. The originality of our approach lies in the fact that the segmentation strategy of the image is based on the distance between the pixels and not on the intensity (colour). The results obtained from test images show that our method is robust and reduces errors due to problems of ambiguity between pixels belonging to areas with low and/or repetitive textures.It also enables us to overcome the problems of occlusions and depth discontinuities.

Quasi-Dense Matching Algorithm for Close-Range Image Combined With Feature Line Constraint

“…On the other hand, geometries are constructed using initial matched points for image segmentation, and seed points to be matched and matching propagation are determined on the basis of the segmented region. Commonly used geometries are Voronoi diagrams [18]- [20] and Delaunay triangulation [21]- [23]. Li et al [24] utilized the Voronoi diagram for constraint propagation to achieve dense matching.…”