Abstract
The technology of 3D displays has made impressive progress as much as technological advance in display, its wide application including entertainment, science, education, and military. However, the technologies of stereoscopic displays abound in the current industry, but they still fail to enter the mass market due to the numerous problems in terms of image quality of stereoscopic displays and asthenopia to viewers. “System crosstalk” is the main cause of 3D image quality and asthenopia, causing the viewers’ mental and physical loading. In addition, depth perception cues have impact on viewers’ subjective evaluation toward image quality as well.In this study, a human factors study on the two-view auto-stereoscopic display was conducted, investigating the relation between system crosstalk and perceived image quality. This study is based on mirror-type stereoscopic display. The independent variables were system crosstalk, shadow, luminosity contrast ratio, and polarity while the dependent variables were subjective evaluation toward image quality as well as objective and subjective evaluation toward asthenopia. The participants were engineers researching on stereoscopic displays and graduate students who never experienced stereoscopic displays. The difference between these two groups was also examined in this study.The experimental results confirm that the main effects of system crosstalk and shadow on perceived stereo image quality is statistically significant. Perceived stereo image quality declines as system crosstalk of images increases. The threshold of system crosstalk for two-view auto-stereoscopic display is recommended to be 8%. In addition, the image showing with shadow elevates stereo image quality. Furthermore, the interaction effect of shadow and luminosity contrast ratio on a sense of depth also attained statistical significance in this study, indicating that brighter background images should take notice of the properly existence of shadow. However, the main effects of luminosity contrast ratio and polarity were not significant; asthenopia caused by stereo images was not due to the insufficient experimental time of watching stereo images, meaning that ten minutes may be too short to cause asthenopia to viewers.