11 Environment and ambient illumination Steve Marschner CS5625 - - PowerPoint PPT Presentation

▶
11 environment and ambient illumination
SMART_READER_LITE
LIVE PREVIEW

11 Environment and ambient illumination Steve Marschner CS5625 - - PowerPoint PPT Presentation

Feb 19, 2023 267 likes •893 views

11 Environment and ambient illumination Steve Marschner CS5625 Spring 2020 Real-time environment illumination Now incident radiance field is stored in a texture, not defined by a polygon this makes it easy to do mirror reflections: single

slide-1
SLIDE 1

11 Environment and ambient illumination

Steve Marschner CS5625 Spring 2020

slide-2
SLIDE 2

Real-time environment illumination

Now incident radiance field is stored in a texture, not defined by a polygon

  • this makes it easy to do mirror reflections: single cubemap lookup

Two special kinds of BRDFs are convenient

  • diffuse BRDFs: irradiance depends only on surface normal (not v)


…leads to irradiance environment maps

  • rotationally symmetric lobes (e.g. Phong): it’s a convolution of the environment map


…leads to prefiltered environment maps

slide-3
SLIDE 3

Standard reflection map

For mirror-specular surface, the illumination integral reduces to This is a function only of r

  • so store it in a cubemap and look up using r(v, n).

Lr(v) = Li(r(v, n))

<latexit sha1_base64="lzHQhec/0+3wgVOSm7sGDJyI+ec=">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</latexit>
slide-4
SLIDE 4

Irradiance environment map

For diffuse surface, illumination integral reduces to This is a function only of n

  • so store it in a cubemap and look up using n.

Lr = R π Z

Ω

Li(l) (n · l) dl

<latexit sha1_base64="IaqHgxQwXRGt6294Fawpmh2WdCM=">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</latexit>
slide-5
SLIDE 5

Irradiance environment map

environment map irradiance map

Gary King in GPU Gems 2 http://http.developer.nvidia.com/GPUGems2/gpugems2_chapter10.html

slide-6
SLIDE 6

Irradiance map illumination

McGuire et al. HPG ’11 10.1145/2018323.2018327

slide-7
SLIDE 7

Irradiance map illumination

McGuire et al. HPG ’11 10.1145/2018323.2018327

slide-8
SLIDE 8

Prefiltered environment map

For a general specular surface, approximate the BRDF by a function

  • where g defines a lobe of width σ that is symmetric around r

The illumination integral is This depends only on r and the scalar σ

  • so store it in an array of cubemaps indexed by σ, and look up using r

fr(v, l) (n · l) ≈ gσ(l · r)

<latexit sha1_base64="qHRkES0l4m/bzqAshvDx3MfcZjw=">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</latexit>

Lr(v) ≈ Z

Ω

gσ(l · r) Li(l) dl

<latexit sha1_base64="lIC2K78XcxySqv6DyL9lSW0PkUI=">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</latexit>
slide-9
SLIDE 9

Irradiance environment mapping

Akenine-Möller et al. RTR 3e

prefiltered map for glossy surface environment map for specular surface prefiltered map for diffuse surface

slide-10
SLIDE 10

Prefiltered environment map

environment map prefiltered for Phong

Gary King in GPU Gems 2 http://http.developer.nvidia.com/GPUGems2/gpugems2_chapter10.html

slide-11
SLIDE 11

blitzcode.net

slide-12
SLIDE 12

Prefiltered map variants

Many approaches to approximating the BRDF with symmetric lobes

  • classic is to use a single lobe, loses stretching at grazing angles
  • many techniques use multiple samples, similar to anisotropic texture filtering
  • modern thought process is to treat the whole precomputation as approximating more accurate

BRDF lobes using multiple samples from arbitrary maps stored in mipmaps

slide-13
SLIDE 13

Manson & Sloan 2016

slide-14
SLIDE 14

Shadow baking

Rendering with no shadows, darker diffuse floor Floor shaded with irradiance from shadow texture Irradiance texture computed using rectangular light

slide-15
SLIDE 15

a convex diffuse object in a constant-radiance environment

slide-16
SLIDE 16

a non-convex diffuse scene under constant-radiance illumination

slide-17
SLIDE 17

a non-convex diffuse object in a constant-radiance environment

slide-18
SLIDE 18

a non-convex diffuse scene under constant-radiance illumination

slide-19
SLIDE 19
  • bject in a constant-radiance environment with no shadowing
slide-20
SLIDE 20

Akenine-Möller et al. RTR 3e

slide-21
SLIDE 21

slide courtesy of Kavita Bala, Cornell University

AO Maps

slide-22
SLIDE 22

Ray traced vertex AO

Kavan et al. EGSR 2011 ambient occlusion sampled at vertices, interpolated as vertex color ambient occlusion sampled inside triangles, vertex values fit to samples, interpolated as vertex color ambient occlusion computed at each pixel (ground truth)

slide-23
SLIDE 23

NVIDIA OptiX implementation images

https://developer.nvidia.com/optix-prime-baking-sample

slide-24
SLIDE 24

NVIDIA OptiX implementation images

https://developer.nvidia.com/optix-prime-baking-sample

slide-25
SLIDE 25

NVIDIA OptiX implementation images

https://developer.nvidia.com/optix-prime-baking-sample

slide-26
SLIDE 26

slide courtesy of Kavita Bala, Cornell University

EnvMap

slide-27
SLIDE 27

slide courtesy of Kavita Bala, Cornell University

AO

slide-28
SLIDE 28

slide courtesy of Kavita Bala, Cornell University

Total

slide-29
SLIDE 29

Akenine-Möller et al. RTR 3e

slide-30
SLIDE 30

slide courtesy of Kavita Bala, Cornell University

Ambient Occlusion: Improvement

  • At each point find

– Fraction of hemisphere that is occluded – Also, average unoccluded direction B

(bent normal) Use B for lighting (see later)

slide-31
SLIDE 31

slide courtesy of Kavita Bala, Cornell University

What about B?

  • The unoccluded direction gives an idea of

where the main illumination is coming from

slide-32
SLIDE 32

slide courtesy of Kavita Bala, Cornell University

SSAO

  • Restrict the hemisphere

– Why? Think of AO in box

  • Typically add a drop-off as you get to the

hemisphere boundary

slide-33
SLIDE 33

slide courtesy of Kavita Bala, Cornell University

Crytek for Crisis: SSAO

  • Take z-buffer
  • Consider sphere around a point p

– Distribute samples – Project to screen space and compare to z buffer

slide-34
SLIDE 34

Martin Mittring, Crysis GmBH http://crytek.com/cryengine/presentations/finding-next-gen-cryengine--2

slide-35
SLIDE 35

Martin Mittring, Crysis GmBH http://crytek.com/cryengine/presentations/finding-next-gen-cryengine--2

slide-36
SLIDE 36

Martin Mittring, Crysis GmBH http://crytek.com/cryengine/presentations/finding-next-gen-cryengine--2

slide-37
SLIDE 37

Hemisphere vs. Sphere

John Chapman http://john-chapman-graphics.blogspot.co.uk/2013/01/ssao-tutorial.html

slide-38
SLIDE 38

Irradiance map + SSAO

McGuire et al. HPG ’11 10.1145/2018323.2018327

slide-39
SLIDE 39

Irradiance map + SSAO

McGuire et al. HPG ’11 10.1145/2018323.2018327

slide-40
SLIDE 40

Irradiance map + SSAO

McGuire et al. HPG ’11 10.1145/2018323.2018327

slide-41
SLIDE 41

Irradiance map + SSAO

McGuire et al. HPG ’11 10.1145/2018323.2018327

slide-42
SLIDE 42

slide courtesy of Kavita Bala, Cornell University

slide-43
SLIDE 43

slide courtesy of Kavita Bala, Cornell University

slide-44
SLIDE 44

slide courtesy of Kavita Bala, Cornell University

slide-45
SLIDE 45

slide courtesy of Kavita Bala, Cornell University

slide-46
SLIDE 46

slide courtesy of Kavita Bala, Cornell University

slide-47
SLIDE 47

slide by Frédo Durand, MIT

Denoising from 1 image

  • We can’t take average over

multiple images Noisy input

slide-48
SLIDE 48

slide by Frédo Durand, MIT

Denoising from 1 image

  • We can’t take average over

multiple images

  • Idea 1: take a spatial

average

  • Most pixels have roughly teh

same color as their neighbor

  • Noise looks high frequency =>

do a low pass

  • Here: Gaussian blur

Noisy input

slide-49
SLIDE 49

slide by Frédo Durand, MIT

Gaussian blur

  • Noise is mostly gone
  • But image is blurry
  • duh!

After Gaussian blur

slide-50
SLIDE 50

slide by Frédo Durand, MIT

Gaussian blur

  • Noise is mostly gone
  • But image is blurry
  • duh!
  • Question: how to blur/

smooth/abstract image, but without destroying important features? After Gaussian blur

adapted from slide by Frédo Durand, MIT

slide-51
SLIDE 51

slide by Frédo Durand, MIT

Bilateral filter

  • [Tomasi and Manduci 1998]

–http://www.cse.ucsc.edu/~manduchi/Papers/ICCV98.pdf

  • Developed for denoising
  • Related to

–SUSAN filter [Smith and Brady 95]

http://citeseer.ist.psu.edu/smith95susan.html

–Digital-TV [Chan, Osher and Chen 2001]

http://citeseer.ist.psu.edu/chan01digital.html

–sigma filter http://www.geogr.ku.dk/CHIPS/Manual/f187.htm

  • Full survey: http://people.csail.mit.edu/sparis/publi/2009/

fntcgv/Paris_09_Bilateral_filtering.pdf

slide-52
SLIDE 52

slide by Frédo Durand, MIT

Start with Gaussian filtering

  • Here, input is a step function + noise
  • utput

input

J

=

f

⊗

I

slide-53
SLIDE 53

slide by Frédo Durand, MIT

Gaussian filter as weighted average

  • Weight of ξ depends on distance to x
  • utput

input

ξ

∑

f (x,ξ)

I(ξ)

ξ x x ξ

J(x) =

slide-54
SLIDE 54

slide by Frédo Durand, MIT

The problem of edges

  • Here, “pollutes” our estimate J(x)
  • It is too different
  • utput

input

x ξ Ι(ξ) I(x) Ι(ξ)

ξ

∑

f (x,ξ)

I(ξ)

J(x) =

slide-55
SLIDE 55

slide by Frédo Durand, MIT

Principle of Bilateral filtering

[Tomasi and Manduchi 1998]

  • Penalty g on the intensity difference
  • utput

input

J(x) =

1 k(x)

ξ

∑ f (x,ξ)

g(I(ξ) − I(x)) I(ξ)

x Ι(ξ) I(x)

slide-56
SLIDE 56

slide by Frédo Durand, MIT

Bilateral filtering

[Tomasi and Manduchi 1998]

  • Spatial Gaussian f
  • utput

input

J(x) =

1 k(x)

ξ

∑ f (x,ξ)g(I(ξ) − I(x)) I(ξ)

x ξ x

slide-57
SLIDE 57

slide by Frédo Durand, MIT

Bilateral filtering

[Tomasi and Manduchi 1998]

  • Spatial Gaussian f
  • Gaussian g on the intensity difference
  • utput

input

J(x) =

1 k(x)

ξ

∑ f (x,ξ) g(I(ξ) − I(x))I(ξ)

x Ι(ξ) I(x)

slide-58
SLIDE 58

slide by Frédo Durand, MIT

Normalization factor

[Tomasi and Manduchi 1998]

  • k(x)=
  • utput

input

J(x) =

1 k(x)

ξ

∑ f (x,ξ)

g(I(ξ) − I(x)) I(ξ)

ξ

∑ f (x,ξ)

g(I(ξ) − I(x))

slide-59
SLIDE 59

slide by Frédo Durand, MIT

Bilateral filtering is non-linear

[Tomasi and Manduchi 1998]

  • The weights are different for each output pixel
  • utput

input

slide-60
SLIDE 60

Cornell CS6640 Fall 2012

Effects of bilateral filter

55

size of domain filter size of range filter

[Tomasi & Manduchi 1998]

slide-61
SLIDE 61

slide by Frédo Durand, MIT

Bilateral filter

Noisy input After gaussian blur

adapted from slide by Frédo Durand, MIT

slide-62
SLIDE 62

slide by Frédo Durand, MIT

Bilateral filter

Noisy input After bilateral filter