Evaluation of Discomfort Degree Estimation System with Pupil Variation in Partial 3D Images

The purpose of this paper was to examine whether the changes in pupil diameter can refrect on the degree of discomfort by various levels of partial 3D images as well as other validated characteristics. Moreover, we discuss the effectiveness of the systems while guiding visual attention by partial 3D images. Images chosen from IAPS (International Affective Picture System) were used to make 3D images. Power spectrum ratio of the pupil variation, called S/C value, generated by stimuli images to those by control images was calculated. The relationship between VAS scores for the impression regarding projected images and the S/C values was set to the major concern for this study. As a result, the average S/C values in 2D neutral images ranged 0.634 to 1.318, whereas the average S/C values in partial 3D neutral images ranged 0.412 to 1.552. VAS scores in 2D neutral images ranged 3.6 to 8.5 and that in partial 3D neutral images ranged 1.2 to 7.4. Moreover, correlation coefficients between VAS scores and S/C values in 2D neutral images was 0.116 and those in partial 3D neutral images was −0.114. In partial 3D images, this negative correlation coefficient was found in consistent with the previous study, whereas the correlation coefficients for both images were relatively low. It was suggested that S/C value was likely to use as a candidate for discomfort measure with a modification of collection technique for VAS scores during the experiment.