Stereo Vision Enabling Precise Border Localization Within a Scanline Optimization Framework

Stefano Mattoccia, Federico Tombari, Luigi di Stefano

Abstract: A novel algorithm for obtaining accurate dense disparity measurements and precise border localization from stereo pairs is proposed. The algorithm embodies a very effective variable support approach based on segmentation within a Scanline Optimization framework. The use of a variable support allows for precisely retrieving depth discontinuities while smooth surfaces are well recovered thanks to the minimization of a global function along multiple scanlines. Border localization is further enhanced by symmetrically enforcing the geometry of the scene along depth discontinuities. Experimental results show a significant accuracy improvement with respect to comparable stereo matching approaches.