SLIDE 1 Page ‹#›
What is NPR?
Amy Gooch Amy Gooch Amy Gooch
Define Photorealistic Rendering
Photo:
- Comes from the Greek phos, meaning light
- r produced by light
Photo: Photo:
Comes from the Greek phos phos, meaning light , meaning light
- r produced by light
- r produced by light
SLIDE 2 Page ‹#›
Define Photorealistic Rendering
Realistic:
–fidelity in art and literature to nature or to real life and to accurate representation without idealization
Realistic: Realistic:
Dictionary definition: – –fidelity in art and literature to nature or to fidelity in art and literature to nature or to real life and to accurate representation real life and to accurate representation without idealization without idealization
Rendering:
- Convert a representation of a virtual scene
into an image for viewing
Rendering: Rendering:
- Convert a representation of a virtual scene
Convert a representation of a virtual scene into an image for viewing into an image for viewing
Understand Photorealistic Rendering
- Light, material properties, reflections &
refraction
- Radiosity, ray tracing, & image-based
rendering
- Light, material properties, reflections &
Light, material properties, reflections & refraction refraction
Radiosity, ray tracing, & image-based , ray tracing, & image-based rendering rendering
SLIDE 3
Page ‹#›
Example: Art of Jan Vermeer Example: Photorealistic Rendering
From Lightscape From From Lightscape Lightscape
SLIDE 4 Page ‹#›
Define Non-Photorealistic Rendering
Descriptors:
- expressive
- artistic
- painterly
- interpretative
Descriptors: Descriptors:
expressive
artistic
painterly
interpretative
Approach
Photorealistic rendering:
- Simulation, usually physically based
Non-photorealistic rendering:
- Stylization, driven by human perception
- Brings together art and science
- Concentrates less on the process and more
- n the results
Photorealistic rendering: Photorealistic rendering:
- Simulation, usually physically based
Simulation, usually physically based
Non-photorealistic rendering: Non-photorealistic rendering:
- Stylization, driven by human perception
Stylization, driven by human perception
- Brings together art and science
Brings together art and science
- Concentrates less on the process and more
Concentrates less on the process and more
- n the results
- n the results
SLIDE 5 Page ‹#›
Characteristic
Photorealistic rendering:
Non-photorealistic rendering:
Photorealistic rendering: Photorealistic rendering:
Objective
Non-photorealistic rendering: Non-photorealistic rendering:
Subjective
Influences
Photorealistic rendering:
- Simulation of physical processes
Non-photorealistic rendering:
- Artistic processes
- Perceptual-based
Photorealistic rendering: Photorealistic rendering:
- Simulation of physical processes
Simulation of physical processes
Non-photorealistic rendering: Non-photorealistic rendering:
Artistic processes
Perceptual-based
SLIDE 6 Page ‹#›
Level of detail
Photorealistic rendering:
- High level of detail, hard to neglect detail
Non-photorealistic rendering:
- Runs the range
- Adapt level of detail across the image to
focus the viewers attention
Photorealistic rendering: Photorealistic rendering:
- High level of detail, hard to neglect detail
High level of detail, hard to neglect detail
Non-photorealistic rendering: Non-photorealistic rendering:
Runs the range
- Adapt level of detail across the image to
Adapt level of detail across the image to focus the viewers attention focus the viewers attention
Level of Abstraction
From Understanding Comics, by Scott McCloud, 1993 From Understanding Comics, by Scott From Understanding Comics, by Scott McCloud McCloud, 1993 , 1993
SLIDE 7
Page ‹#›
Illustrators Use of Lines
From Technical Illustration by Judy Martin From From Technical Illustration Technical Illustration by Judy Martin by Judy Martin
Wire frame versus Edge Lines
Are Real Edges Sufficient For Object Recognition? Are Real Edges Sufficient For Object Recognition? Sanocki et al. Sanocki et al.
SLIDE 8
Page ‹#›
Silhouettes
Submitted to The Journal of Graphics Tools: Submitted to The Journal of Graphics Tools: Evaluating Silhouette Algorithms for Polygonal Models, Hartner et al. Evaluating Silhouette Algorithms for Polygonal Models, Hartner et al. A Top-Down Approach to Normal Cone Hierarchies, Hartner et al. A Top-Down Approach to Normal Cone Hierarchies, Hartner et al. Evaluating Object Space Methods for Silhouette Rendering, Beddes et al. Evaluating Object Space Methods for Silhouette Rendering, Beddes et al.
Creases
SLIDE 9
Page ‹#›
Surface boundaries Silhouettes in Black, Creases in White
Gooch et al., I3D 1999 Gooch et al., Gooch et al., I3D 1999 I3D 1999
SLIDE 10
Page ‹#›
Artistic Silhouettes
Gooch et al., I3D 1999 Gooch et al., Gooch et al., I3D 1999 I3D 1999
Shading Used by Artists
Complementary Shading Complementary Shading Complementary Shading Final image Final image Final image
The Book of Color, by Jose Parramon, 1993 The Book of Color, The Book of Color, by Jose Parramon, 1993 by Jose Parramon, 1993
SLIDE 11
Page ‹#›
Tone Shading on a Gray Model Tone Shading
Tone Shading Tone Shading Tone Shading Phong Shading Phong Shading Phong Shading Gooch et al., ACM Siggraph 1998 Gooch et al., Gooch et al., ACM Siggraph 1998 ACM Siggraph 1998
SLIDE 12
Page ‹#›
Treasure Planet Material Properties
Illustration Illustration Illustration Photograph Photograph Photograph
SLIDE 13
Page ‹#›
Imitating Material Properties
Phong shaded Phong shaded Phong shaded Metal shaded Metal shaded Metal shaded
Illustrators Use of Shadows
SLIDE 14
Page ‹#›
Soft Colored Shadow
Gooch et al., Gooch et al., ACM I3D 1999 ACM I3D 1999
Mayan Temple Example
Potter et al. Potter et al. Potter et al.
SLIDE 15 Page ‹#›
Motivation for Computer Scientists
Provide tools for :
Provide tools for : Provide tools for :
artists
non-artists
Categorizing NPR
Hard to categorize NPR
–pen-an-ink –watercolor –oil
Hard to categorize NPR Hard to categorize NPR
Natural media emulation – –pen-an-ink pen-an-ink – –watercolor watercolor – –oil
Curtis et al. 1997 Curtis et al. Curtis et al. 1997
1997
Hertzmann 1997 Hertzmann Hertzmann 1997
1997
Winkenbach et al. 1994 Winkenbach Winkenbach et al. et al. 1994
1994
SLIDE 16 Page ‹#›
Line and Stipple Drawing
Secord et al. 2002 Secord Secord et al. 2002 et al. 2002
Deussen et al. Deussen Deussen et al. et al. Way et al. 2002 Way et al. 2002 Way et al. 2002 Sousa et al. 2003 Sousa et al. 2003 Sousa et al. 2003
Hatching
Webb et al. 2002 Webb et al. 2002 Webb et al. 2002 Praun et al. 2001 Praun Praun et al. 2001 et al. 2001 Hertzmann 2000 Hertzmann Hertzmann 2000 2000
SLIDE 17 Page ‹#›
Cartoon Rendering
Markosian et al. Markosian Markosian et al. et al. Kaplan et al. Kaplan et al. Kaplan et al. Lake et al. Lake et al. Lake et al.
Simulating materials
Wax Crayons, Rudolf et al. 2003 Wax Crayons, Rudolf et al. 2003 Wax Crayons, Rudolf et al. 2003
Rudolf et al. 2003 Rudolf et al. 2003 Rudolf et al. 2003
SLIDE 18 Page ‹#›
Embossing
Sourin 2001 Sourin Sourin 2001 2001
Categorizing NPR
- Image and video enhancement
- Image and video enhancement
Image and video enhancement
Before Before Before After After After From www.revisionfx.com From www. From www.revisionfx revisionfx.com .com
SLIDE 19 Page ‹#›
Categorizing NPR
- Image and video enhancement
- Image and video enhancement
Image and video enhancement
From www.revisionfx.com From www. From www.revisionfx revisionfx.com .com
Categorizing NPR
- User interface, like a paint program (2D & 3D)
- User interface, like a paint program (2D & 3D)
User interface, like a paint program (2D & 3D)
Haeberli 1990 Haeberli Haeberli 1990 1990 Salisbury et al. 1997 Salisbury et al. 1997 Salisbury et al. 1997 Cohen 2000 Cohen 2000 Cohen 2000 Gooch et al. 2002 Gooch et al. 2002 Gooch et al. 2002
SLIDE 20 Page ‹#›
Categorizing NPR
- Process applied to geometric models
- Process applied to geometric models
Process applied to geometric models
Meier 1996 Meier Meier 1996
1996
Saito et al. 1990 Saito et al. Saito et al. 1990
1990
Dooley et al. 1990 Dooley et al. Dooley et al. 1990
1990
Categorizing NPR
- Process applied to geometric models
- Process applied to geometric models
Process applied to geometric models
Gooch et al. 1998 Gooch et al. 1998 Gooch et al. 1998
SLIDE 21 Page ‹#›
Creating Models
http://www-ui.is.s.u-tokyo.ac.jp/~takeo/java/smoothteddy/ Takeo Igarashi Takeo Igarashi
Sketch From Brown U. Sketch From Brown U. Sketch From Brown U.
What NPR can & can't/shouldn't do
- It is difficult to imitate the decisions of a real
artist
- Can we give computer programs
inventiveness?
- Can programs be expressive in them
selves, or do we depend on the user?
- It is difficult to imitate the decisions of a real
It is difficult to imitate the decisions of a real artist artist
- Can we give computer programs
Can we give computer programs inventiveness? inventiveness?
- Can programs be expressive in them
Can programs be expressive in them selves, or do we depend on the user? selves, or do we depend on the user?
SLIDE 22 Page ‹#›
It’s not all automatic
You can't just apply NPR to any image or scene & get a work of art Think about:
- Subject matter
- Scene composition
- Conveying emotion or purpose
You can't just apply NPR to any image or scene You can't just apply NPR to any image or scene & get a work of art & get a work of art Think about: Think about:
Subject matter
Scene composition
- Conveying emotion or purpose
Conveying emotion or purpose
Benefits of NPR
- Communication
- Change user focus
- Eliminate unnecessary details
- Create details not seen in the geometric
model
Communication
Change user focus
- Eliminate unnecessary details
Eliminate unnecessary details
- Create details not seen in the geometric
Create details not seen in the geometric model model
SLIDE 23
Page ‹#›
