17
1 Monte Carlo Global Illumination Brandon Lloyd COMP 238 December 16, 2002

1 Monte Carlo Global Illumination Brandon Lloyd COMP 238 December 16, 2002

  • View
    217

  • Download
    0

Embed Size (px)

Citation preview

Page 1: 1 Monte Carlo Global Illumination Brandon Lloyd COMP 238 December 16, 2002

1

Monte Carlo Global Illumination

Brandon Lloyd

COMP 238December 16, 2002

Page 2: 1 Monte Carlo Global Illumination Brandon Lloyd COMP 238 December 16, 2002

2

Monte Carlo Method

• Advantages– Good for integrals of high dimension– All you need is point samples– Allows for arbitrary number of samples

• Disadvantages– Susceptible to noise (caused by high frequencies in

the integrand)– Slow convergence where N is the number

of samples)( NO

Page 3: 1 Monte Carlo Global Illumination Brandon Lloyd COMP 238 December 16, 2002

3

Monte Carlo Method

• The expected value of a function f according to a pdf p:

• Can be approximated with a discrete number of samples xi ~ p (converges as N)

dxxpxfxfE )()()]([

N

iixfN

xfE1

)(1

)]([

Page 4: 1 Monte Carlo Global Illumination Brandon Lloyd COMP 238 December 16, 2002

4

Monte Carlo Method

• … but we are interested in the integral of an arbitrary function f.

N

i i

i

xp

xf

Nxf

1 )(

)(1)(

N

iii xf

Nxpxf

1

)(1

)()(

Page 5: 1 Monte Carlo Global Illumination Brandon Lloyd COMP 238 December 16, 2002

5

Importance Sampling

• We can use any distribution p that is non-zero over the domain

• The distribution affects variance • The more closely p matches f the less variance

you will have.• If p = f then you get the right answer with one

sample! But that requires we know f.

N

i i

i

xp

xf

Nxf

1 )(

)(1)(

Page 6: 1 Monte Carlo Global Illumination Brandon Lloyd COMP 238 December 16, 2002

6

Importance Sampling

• Directional formulation of the rendering equation:

• We don’t know Li . We can sample according to: f, cos , or f cos

iiiioiro dLxfxL cos)(),(),(

Page 7: 1 Monte Carlo Global Illumination Brandon Lloyd COMP 238 December 16, 2002

7

Importance Sampling

• Point formulation of the rendering equation:

• A bit more complicated. Usually just generate points on the surfaces.

)(coscos

),(),(),,(),( 2 xdAxx

xxVxxLxxxfxxLS

ir

Page 8: 1 Monte Carlo Global Illumination Brandon Lloyd COMP 238 December 16, 2002

8

Generating Samples

• We can easily generate a uniform random variable U.

• Use the Inversion Method to transform U to X ~ p.

– Create the CDF of p

– Use the inverse of P to transform U.

x

dsspxP )()(

)(1 UPX

Page 9: 1 Monte Carlo Global Illumination Brandon Lloyd COMP 238 December 16, 2002

9

Example: Diffuse BRDF

• Choose

iiiio dLxL cos)(),(

2

0 0

sincos)(2

iiiiii ddL

sincos),( p

Page 10: 1 Monte Carlo Global Illumination Brandon Lloyd COMP 238 December 16, 2002

10

Example: Diffuse BRDF

• p is separable so we treat each dimension independently

• Invert by solving for u0 = P and u1 = P

22

1

cossincos

0

0

2

dP

dP

)2),(arccos(),( 21 uu

Page 11: 1 Monte Carlo Global Illumination Brandon Lloyd COMP 238 December 16, 2002

11

Example: Diffuse BRDF

• Final Estimator

• The Global Illumination Compendium [Dutre 2001] contains transformations for a number of useful pdfs that arise in global illumination problems

)(1

),( iio LN

xL

Page 12: 1 Monte Carlo Global Illumination Brandon Lloyd COMP 238 December 16, 2002

12

Tranforming the Distribution

• The distribution is created in a canonical space but we need to have it about the surface normal.

ZN

Page 13: 1 Monte Carlo Global Illumination Brandon Lloyd COMP 238 December 16, 2002

13

Tranforming the Distribution

• Obvious method. Create a coordinate frame by picking arbitrary S. T = ||NxS|| S=||TxN||

• Can be done more cheaply [Hughes99]• If the distribution is isotropic then reflect about

the half-way vector

Z

H

N

Page 14: 1 Monte Carlo Global Illumination Brandon Lloyd COMP 238 December 16, 2002

14

Results

Test Scene

Page 15: 1 Monte Carlo Global Illumination Brandon Lloyd COMP 238 December 16, 2002

15

BRDF sampling Area sampling

Path tracing (combined sampling)

Multiple Importancesampling

Page 16: 1 Monte Carlo Global Illumination Brandon Lloyd COMP 238 December 16, 2002

16

Path tracing

Multiple ImportanceSampling

Multiple ImportanceSampling

Bias!

Page 17: 1 Monte Carlo Global Illumination Brandon Lloyd COMP 238 December 16, 2002

17

References

[Hughes99] John F. Hughes and Tomas Möller, “Building an Orthonormal Basis from a Unit Vector'' Journal of Graphics Tools, vol. 4, no. 4, pp. 33-35, 1999.

[Dutre01] Phillip Dutre, Global Illumination Compendium, http://www.graphics.cornell.edu/~phil/GI/, 2001