Image Rotation Rectification in steroscopic 3D on multi-core architectures

Ivan Velciov, Cormac Brick, Marius Predut, Valentin Muresan
In this paper, a rectification procedure necessary for obtaining a quality stereoscopic 3D effect, is presented. Rotation rectification is necessary due to misalignment of the pair of image sensors. In particular when the misalignment angle is, greater than 0.7 degrees, the perceived quality of the final 3D stereoscopic image is degraded. This paper offers a low-power, mobile, multi-core solution.