Mikhail Sizintsev, Richard P. Wildes
Image and Vision Computing, Volume 28, Issue 3, March 2010
This paper presents methods for efficient recovery of accurate binocular disparity estimates in the vicinity of 3D surface discontinuities. Of particular concern are methods that impact coarse-to-fine, local block-based matching as it forms the basis of the fastest and the most resource efficient stereo computation procedures. A novel coarse-to-fine refinement procedure that adapts match window support across scale to ameliorate corruption of disparity estimates near boundaries is presented. Extensions are included to account for half-occlusions and colour uniformity. Empirical results show that incorporation of these advances in the standard coarse-to-fine, block matching framework reduces disparity errors by more than a factor of two, while performing little extra computation, preserving low complexity and the parallel/pipeline nature of the framework. Moreover, the proposed advances prove to be beneficial for CTF global matchers as well.
Importantly, in the follow-up research work, the proposed high-accuracy and low-complexity algorithm was realized in standard GPU  and FPGA  frameworks to perform dense stereo computation on 640×480 images at rates above 30 Hz. M. Sizintsev, S. Kuthirummal, H. Sawhney, A. Chaudhry, S. Samarasekera and R. Kumar, “GPU Accelerated Realtime Stereo for Augmented Reality ”, In the Proceedings of the 5th International Symposium on 3D Data Processing, Visualization and Transmission (3DPVT), 2010  Eduardo Gudis, Gooitzen van der Wal, Sujit Kuthirummal and Sek Chai, “Multi-Resolution Real-Time Dense Stereo Vision Processing in FPGA”, In the Proceedings on IEEE Symposium on Field-Programmable Custom Computing Machines (FCCM), 2012
Go to Journal