基于样图的曲面纹理合成-NaiwenXie谢乃闻.PPTVIP

The third step, ec is computed by displacing e0 towards or away from the centroid of the cluster Pc. The displacement is controlled by a scalar value f, which is set such that the diagonal of the 2-D reflected ray ends points AABB be approximately equal to the diagonal of the 2-D reflected ray starts poings AABB We use this generated camera to render the impostor. * * Due to the nonlinear mapping induced by the perspective distortion, Pixel space and the 3D homogeneous space are related by a linear transformation. A z-value related function is used to find the intersection place in image place of a point along the segment S-E. / Thus this equation gives the Z coordinate, in camera space, of a point along the segment S-E as a function of the parameter α defined in image space. / * * The? glossy attribute controls light that is reflected in one direction, but slightly scattered. It is used to produce blurry reflections off surfaces that have a slight roughness. A point on a glossy surface does not reflect along a single direction, rather along a solid angle centered at the specularly reflected ray. We do glossy effect by setting the far plane of a cluster camera constructed to render a glossy reflection based on the glossiness level. The reflected rays sample uniformly a cone with apex at the surface point and with an axis defined by the reflected ray generated by the pixel normal. * * Thougout the 3 picture, our method are comparable to reflections rendered by ray tracing. The reflections effect rendered with our method are rendered at a resolution of 512 x 512 with uniform 2x2 super-sampling. We use NVIDIA’s Optix ray tracer to compare the reslut. We compare the result by minute our picture value with the ray-tracing picture. Here is the two results. Only a small number of pixels has flaws. * * One reason for the difference is the large angle between reflected rays at the reflector edges, which leads to minification errors. Another reason for