Polygon-based computer-generated holography: Anti-aliased texture rendering via self-similar segmentation

In textured polygon-computer-generated holography based on self-similar segmentation, texture can represent the details of an object without increasing the number of meshes. Texture mapping is usually realized by sampling texture values, which produces aliasing artifacts at abrupt changes in texture values when the subdivision degree M is small with insufficient number of samples. Although this can be mitigated by increasing M, it causes a dramatic increase in computational complexity. To balance the computational effort with the visual quality, in this paper, we propose an effective anti-aliasing texture rendering method that utilizes sub-triangle vertices to sample the texture and linearly interpolates the sampled texture values to assign inhomogeneous magnitudes to the primitive sub-triangles. Our method can mitigate aliasing artifacts to improve reconstruction quality with acceptable calculation time. Numerical and optical reconstructions demonstrate the effectiveness of the scheme.