Visual discomfort prediction for stereo contents
Shan He, Tao Zhang, Didier Doyen · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2011
The current renaissance of 3D movies has drawn more and more attention from the audience. Three-dimensional television (3DTV) has been expected to be the next advance in television. Studies have shown that different people have different comfort range of depth in a 3D content, especially in 3DTV scenario, wherein much smaller screen sizes and viewing distances in home setup than in theater put more restrictions on the 3D content fed into the 3DTV. As a result, the version of the 3D content sent to home will not satisfy all the people in one family. In this paper, we try to solve this problem by providing a prediction of viewing discomfort of certain input content by certain viewer. Our method is based on the Disparity Discomfort Profile (DDP) built through subjective test for each viewer. The input content is analyzed by studying its disparity distribution. The prediction of discomfort is performed by matching the disparity distribution with the viewer's DDP. Then a mechanism to allow the viewers to adjust the depth range according to their visual comfort profile or viewing preference is used to minimize the discomfort. Experiments show promising results of the proposed method.