Computer Graphics Normal, Texture, Supersampling Exercises - PowerPoint PPT Presentation

1 / 6
About This Presentation
Title:

Computer Graphics Normal, Texture, Supersampling Exercises

Description:

Normal, Texture, Supersampling Exercises - Marcus Magnor ... Invertible if points are non-collinear. Computer Graphics WS05/06 Supersampling Exercises ... – PowerPoint PPT presentation

Number of Views:43
Avg rating:3.0/5.0
Slides: 7
Provided by: marcus45
Category:

less

Transcript and Presenter's Notes

Title: Computer Graphics Normal, Texture, Supersampling Exercises


1
Computer Graphics- Normal, Texture,
Supersampling Exercises -
  • Marcus Magnor

2
Vertex Normals
  • Normal vector per vertex
  • E.g., averaged orientations of adjoining
    triangles
  • Normalized
  • Normal interpolation in triangle
  • Hitpoint position in barycentric coordinates
    (u,v,1-u-v)
  • Barycentric interpolation cf. lecture
    raytrace1, slide 36
  • Re-normalize interpolated normal vector

3
Surface Parameterization
  • Some objects have a natural parameterization
  • Sphere spherical coordinates (?,f) (2p u, p v)
  • Cylinder cylindrical coordinates (?,z) (2p u,
    H v)
  • Parametric surfaces (such as B-spline or Bezier
    surfaces)
  • Parametrization less obvious for
  • Polygons
  • Implicit surfaces

4
Triangle Parameterization
  • Piecewise planar object surface patches
  • To each vertex (xo,yo,zo) assign texture
    coordinates (u,v)
  • Viewing projection (xo,yo,zo) ? (x,y)
  • Yields warping operation (u,v)?(x,y)
  • In homogeneous coordinates
  • Transformation coefficients determined by 3 pairs
    (u,v)?(x,y)
  • Invertible if points are non-collinear

5
Triangle Parameterization II
  • Inverse transform (x,y)?(u,v)
  • Transformation coefficients must be calculated
    for each triangle
  • Scan-line rendering
  • Bilinear interpolation of (u,v,q) in screen
    space
  • Ray tracing
  • Evaluation at each intersection

6
Supersampling
  • Regular
  • Subdivide pixel into sub-areas
  • Send one ray through each sub-areas center
  • Average
  • Random
  • N rays through pixel area at (quasi-)random 2D
    positions
  • average
  • Stratified
  • Combination of former two
  • Subdivide pixel into sub-areas
  • Send one ray through each sub-area at (quasi-)
    random 2D position
  • average
Write a Comment
User Comments (0)
About PowerShow.com