Temporal interlacing and spatial interlacing are the two most common methods of stereoscopic 3D presentation, but each has its own shortcomings. Temporal interlacing is prone to more salient motion artifacts and distortion in perceived depth, while spatial interlacing has a lower effective resolution at the recommended viewing distance. We proposed a novel stereoscopic 3D technique that uses both temporal and spatial interlacing: one subframe shows the left-eye view on odd pixel-rows and the right-eye view on even pixel-rows while the other subframe shows the right-eye view on odd pixel-rows and the left-eye view on odd pixel-rows. It is the same as spatial interlacing in a single subframe, but the assigned view for each pixel-row changes across every subframe. Hence, we termed this approach a hybrid stereoscopic 3D method. Through a series of psychophysical experiments, we verified that the proposed method outperforms both temporal interlacing (in terms of motion artifacts and depth representation) and spatial interlacing (in terms of effective resolution).
Collaborators: Paul Johnson [home], Martin Banks