Performance-improved smart pseudoscopic to orthoscopic conversion for integral imaging by use of lens array shifting technique

The smart pseudoscopic to orthoscopic conversion algorithm is thought to be the most generalized method to solve the pseudoscopic problem in integral imaging. However, accurate pixel mapping can only be achieved for object points near the reference plane, moreover, correct elemental image array may not be possible for large-depth objects with a single reference plane. In this paper, we investigate the influence of spatial sampling rate in pickup process on the performance of the SPOC algorithm, and propose to apply the lens array shifting technique to the algorithm, so that accurate EIAs can be generated with a freely chosen reference plane even for large-depth objects. Experiments are conducted to verify the feasibility of the proposed method.