Anti-Aliasing Techniques Alliased vs. Antialiased 1 Polygon Edges - - PDF document

anti aliasing techniques alliased vs antialiased
SMART_READER_LITE
LIVE PREVIEW

Anti-Aliasing Techniques Alliased vs. Antialiased 1 Polygon Edges - - PDF document

Anti-Aliasing Techniques Alliased vs. Antialiased 1 Polygon Edges Area Sampling 2 Catmull, 1978 Carpenter, 1984 A-Buffer 3 The A-Buffer Idea: approximate continuous filtering by subpixel sampling Summing areas now becomes


slide-1
SLIDE 1

1

Anti-Aliasing Techniques Alliased vs. Antialiased

slide-2
SLIDE 2

2

Polygon Edges Area Sampling

slide-3
SLIDE 3

3

Catmull, 1978 Carpenter, 1984 – A-Buffer

slide-4
SLIDE 4

4

The A-Buffer

  • Idea: approximate continuous filtering

by subpixel sampling

  • Summing areas now becomes simple
slide-5
SLIDE 5

5

Abram & Westover, 1983

  • Abram & Westover, 1983
slide-6
SLIDE 6

6

Abram & Westover, 1983 SuperSampling

Sample at 4x4 subpixels and color the pixel according to the portion of the coverage.

slide-7
SLIDE 7

7

Andreas Schilling

  • Andreas Schilling
slide-8
SLIDE 8

8

Andreas Schilling

12 13 14 15 8 9 10 11 4 5 6 7 1 2 3

  • Catmull
  • s i j

I i x j y dxdy

A i j

( , ) ( , )

( , )

  • s(i,j)

(i,j) A(i,j)

  • Carpenter

s i j I i a n j b n F a n b n

b n a n

( , ) ( , ) ( , )

  • 1

1

n Abram & Westover s i j c i j F

p p p P

( , ) ( , )

  • 
  • Fp
  • P

pcp(i,j)=0 (i,j)pcp(i,j)p cp(i,j)=0 cp2(i,j)=-1 cp1(i,j)=1

  • 0- p cp(i,j)=1
  • pp1,p2
  • Schilling

s i j O i j F a n b n

a b b n a n

( , ) ( , ) ( , )

,

  • 1

1

(i,j)(a,b)Oa,b(i,j)

slide-9
SLIDE 9

9

  • Catmull

Carpenter Abram & Westover Schilling

  • (x,y)F(x,y) = 1
  • Texture Aliasing
  • A single screen space pixel might correspond to

many texels (texture elements):

slide-10
SLIDE 10

10

  • Two special cases:
  • Magnification: No real need in prefiltering.

The main decision is what kind of reconstruction (interpolation) to use.

  • Minification: No real need in
  • reconstruction. The main problem is proper

prefiltering.

slide-11
SLIDE 11

11

Nearest neighbor sampling Filtered Texture:

slide-12
SLIDE 12

12

Texture Pre-Filtering

  • Problem: filtering the texture during

rendering is too slow for interactive performance.

  • Solution: pre-filter the texture in advance

– Summed area tables - gives the average value

  • f each axis-aligned rectangle in texture space

– Mip-maps (tri-linear interpolation) - supported by most of today’s texture mapping hardware

slide-13
SLIDE 13

13

MIP-Maps

  • Precompute a set of prefiltered textures

(essentially an image pyramid).

  • Based on the area of the pre-image of the

pixel:

– Select two “best” resolution levels – Use bilinear interpolation inside each level – Linearly interpolate the results

  • Referred to as trilinear interpolation

Summed Area Tables

  • A 2D table the size of the texture. At each

entry (i,j), store the sum of all texels in the rectangle defined by (0,0) and (i,j).

  • Given any axis aligned rectangle, the sum
  • f all texels is easily obtained from the

summed area table:

C C B B A A D D

D C B A area

  • D

C B A area

slide-14
SLIDE 14

14

slide-15
SLIDE 15

15

Linear interpolation of texture coordinates Perspective-correct texture coordinate interpolation

slide-16
SLIDE 16

16

Resampling a signal

  • Given a sampled (discrete) signal:

– Reconstruct a continuous signal – Warp (apply mapping) – Prefilter warped signal – Sample

  • In practice, some of the above stages are

collapsed into a single convolution