Bilaterally Weighted Patches for Disparity Map Computation

Laura Fernández Julià (1, 2, 3) , Pascal Monasse (1, 2, 3)

1 imagine [Marne-la-Vallée]
2 LIGM - Laboratoire d'Informatique Gaspard-Monge
3 ENPC - École des Ponts ParisTech

Visual correspondence is the key for 3D reconstruction in binocular stereovision. Local methods perform block-matching to compute the disparity, or apparent motion, of pixels between images. The simplest approach computes the distance of patches, usually square windows, and assumes that all pixels in the patch have the same disparity. A prominent artifact of the method is the " foreground fattening effect " near depth discontinuities. In order to find a more appropriate support, Yoon and Kweon introduced the use of weights based on color similarity and spatial distance, analogous to those used in the bilateral filter. This paper presents the theory of this method and the implementation we have developed. Moreover, some variants are discussed and improvements are used in the final implementation. Several examples and tests are presented and the parameters and performance of the method are analyzed.

Domaines: Vision par ordinateur et reconnaissance de formes [cs.CV]

Laura Fernández Julià, Pascal Monasse. Bilaterally Weighted Patches for Disparity Map Computation. Image Processing On Line, 2015, 5, pp.73-89. ⟨10.5201/ipol.2015.123⟩. ⟨hal-01131412⟩