Quality enhancement of the multi-view exhibition system

Spatial Psychovisual Modulation (SPVM) is a newly proposed display technology which makes use of the redundancy in display capacity. Many applications (i.e., information security display and simultaneous dual-subtitle exhibition) have been developed based on SPVM. In the implementation of the SPVM systems, the original images are down-sampled line-by-line due to the property of polarization screens. The sampling process causes aliasing and influence negatively on the image quality. In this paper, we add a Lowpass Filter (LPF) into the original system before the sampling to eliminate the aliasing. We compare the improved algorithm and the direct method with objective image quality assessment (IQA). For both natural and computer generated images are used in SPVM, we tested our algorithm on many different databases. Considering the view distance of the SPVM system, we used the SAST model as the metrics of image quality. As a result, the proposed algorithm improves the IQA benchmark, which corresponds to our observation. The test also works as an application and test bed of the SAST IQA model. All the equipments of the multi-view exhibition system is off-the-shelf so it's cheaper to realize. And the enhancement in quality make the system capable for screens with lower PPI.