Creating and sharing knowledge for telecommunications

Parallel refinement of slanted 3D reconstruction using dense stereo induced from symmetry

Ralha, R. R. ; Falcão, G. ; Amaro, J. ; Mota, V. G. M. ; Antunes, M. A. ; Barreto, J. P. B. ; Nunes, U.

Journal of Real-Time Image Processing Vol. 1, Nº 1, pp. 1 - 19, May, 2016.

ISSN (print): 1861-8200
ISSN (online): 1861-8219

Scimago Journal Ranking: 0,31 (in 2016)

Digital Object Identifier: 10.1007/s11554-016-0592-0

Traditional dense stereo estimation algorithms measure photo-similarity to calculate the disparity between image pairs. SymStereo is a new framework of matching cost functions that measure symmetry to evaluate the possibility of two pixels being a match. This article pro- poses a fully functional real-time parallel 3D reconstruc- tion pipeline that uses dense stereo-based photo-symmetry. The logN variant of SymStereo achieves superior results for images with slanted surfaces, when compared with other algorithms (Antunes and Barreto in Int J Comput Vis 1–22, 2014). This is of particular interest for areas of computer vision such as the processing of datasets for urban scene reconstruction and also for tracking in robotics