(英) |
When humans perceive a 3D object, they visually check the surface of the object and evaluate the shape and condition of the object sensitively. On the other hand, when constructing an objective information system using computer graphics, it is desirable to quantitatively evaluate the texture from the amount of multidimensional information by quantifying some visual information on the object's surface. In order to accurately predict the recognition of the material quality of a 3D CG object, it is necessary to analyze comprehensive multidimensional information about the texture, including color information, contrast, and rotation factors, in addition to luminance information. In this study, we statistically analyzed and evaluated images for contrast and rotation elements, HEVC, and texture information of 3D CG objects using images in the texture perception standard task image data set. |