lecture 8 - massachusetts institute of technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf ·...
TRANSCRIPT
![Page 1: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/1.jpg)
Lecture 8 Textures
![Page 2: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/2.jpg)
What is a texture?
![Page 3: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/3.jpg)
![Page 4: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/4.jpg)
![Page 5: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/5.jpg)
![Page 6: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/6.jpg)
Which textures are we going to talk about in this lecture?
Stationary
Stochastic
![Page 7: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/7.jpg)
When are two textures similar?
All these images are different instances of the same texture We can differentiate between them, but they seem generated by the same process
![Page 8: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/8.jpg)
Texture Analysis
Compare textures and decide if they’re made of the same “stuff”.
True (infinite) texture
ANALYSIS
generated image
input image
“Same” or “different”
![Page 9: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/9.jpg)
Texture Synthesis
Given a finite sample of some texture, the goal is to synthesize other samples from that same texture – The sample needs to be "large enough“
True (infinite) texture
SYNTHESIS
generated image
input image
![Page 10: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/10.jpg)
Let’s get a feeling of the mechanisms for
texture perception
![Page 11: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/11.jpg)
What is special about texture perception?
• Pre-attentive texture discrimination • Perception of sets and summary statistics • Crowding
![Page 12: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/12.jpg)
Pre-attentive texture discrimination
Bela Julesz, "Textons, the Elements of Texture Perception, and their Interactions". Nature 290: 91-97. March, 1981.
![Page 13: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/13.jpg)
Pre-attentive texture discrimination
Bela Julesz, "Textons, the Elements of Texture Perception, and their Interactions". Nature 290: 91-97. March, 1981.
![Page 14: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/14.jpg)
Pre-attentive texture discrimination
Bela Julesz, "Textons, the Elements of Texture Perception, and their Interactions". Nature 290: 91-97. March, 1981.
![Page 15: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/15.jpg)
Pre-attentive texture discrimination
Bela Julesz, "Textons, the Elements of Texture Perception, and their Interactions". Nature 290: 91-97. March, 1981.
This texture pair is pre-attentively indistinguishable. Why?
![Page 16: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/16.jpg)
![Page 17: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/17.jpg)
Crowding
KBW R +
Pelli, D. G., Cavanagh, P., Desimone, R., Tjan, B., & Treisman, A. (2007). Crowding: Including illusory conjunctions, surround suppression, and attention. Journal of Vision, 7(2):i, 1, http://journalofvision.org/7/2/i/
![Page 18: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/18.jpg)
Journal of Vision November 19, 2009 vol. 9 no. 12
![Page 19: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/19.jpg)
Where’s waldo?
![Page 20: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/20.jpg)
![Page 21: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/21.jpg)
Representation of sets
![Page 22: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/22.jpg)
Set Is this element a member of the set?
![Page 23: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/23.jpg)
Ebbinghaus illusion
Attenuated by reducing the set grouping.
The central circle is judged relative to the set properties of the circles surrounding it
![Page 24: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/24.jpg)
Representation
![Page 25: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/25.jpg)
What a model should account for: 1. Biological plausibility: The stages of the model should
be motivated by, and be consistent with, known physiological mechanisms of early vision.
2. Generality: The model should be general enough that it can be tested on any arbitrary gray-scale image.
3. Quantitative match with psychophysical data: The
model should make a quantitative prediction about the salience of the boundary between any two textured regions. Rank ordering of the discriminability of different texture pairs should agree with that measured psychophysically.
From Malik & Perona, 1990
![Page 26: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/26.jpg)
Julesz - Textons
![Page 27: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/27.jpg)
Julesz - Textons Textons: fundamental texture elements. Textons might be represented by features
such as terminators, corners, and intersections within the patterns…
![Page 28: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/28.jpg)
“We note here that simpler, lower-level mechanisms tuned for size may be sufficient to explain this discrimination.”
Observation: the Xs look smaller than the Ls.
![Page 29: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/29.jpg)
Ls 25% larger contrast adjusted to keep mean constant
Ls 25% shorter
![Page 30: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/30.jpg)
![Page 31: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/31.jpg)
image
Squared responses Spatially blurred
Threshold squared, blurred responses, then categorize texture based on those two bits
vertical filter
horizontal filter
![Page 32: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/32.jpg)
![Page 33: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/33.jpg)
Two big families of models
1- Parametric models of filter outputs
2- Example-based non-parametric models
![Page 34: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/34.jpg)
The trivial texture synthesis algorithm
![Page 35: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/35.jpg)
Texture synthesis and representation
Space of all images
Set of equivalent textures
Set of equivalent textures: generated by exactly the same physical process
![Page 36: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/36.jpg)
Texture synthesis and representation
Space of all images
Set of equivalent textures
Set of perceptually equivalent textures
Set of equivalent textures: generated by exactly the same physical process Set of perceptually equivalent textures: “well, they just look the same to me”
![Page 37: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/37.jpg)
If matching the averaged squared filter values is a good way to match a given texture, then maybe matching the entire marginal distribution (eg, the histogram) of a filter’s response would be even better.
Jim Bergen proposed this…
![Page 38: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/38.jpg)
SIGGRAPH 1994
![Page 39: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/39.jpg)
The main idea: it works by ‘kind of’ projecting a random image into the set of equivalent textures
Space of all images
Set of equivalent textures
Set of perceptually equivalent textures
![Page 40: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/40.jpg)
Overview of the algorithm
Two main tools:
1- steerable pyramid
2- matching histograms
![Page 41: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/41.jpg)
1-The steerable pyramid
+
Low-pass residual
![Page 42: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/42.jpg)
1-The steerable pyramid
But why do I want to represent images like this?
![Page 43: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/43.jpg)
1-The steerable pyramid
Argument used by H & B: Statistical measures in the subband representation seem to provide a “distance” between textures that correlates with human perception better than pixel-based representations.
![Page 44: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/44.jpg)
1-The steerable pyramid
In general seems a good idea to have a representation that:
-Preserves all image information (we can go back to the image)
-Provides more independent channels of information than pixel values (we can mess with each band independently)
But all this is just indirectly related to the texture synthesis task. But let assume is good enough…
![Page 45: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/45.jpg)
Input texture
Steerable pyr
1-The steerable pyramid
![Page 46: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/46.jpg)
Overview of the algorithm
Two main tools:
1- steerable pyramid
2- matching histograms
![Page 47: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/47.jpg)
2-Matching histograms
9% of pixels have an intensity value within the range[0.37, 0.41]
75% of pixels have an intensity val smaller than 0.5
5% of pixels have an intensity value within the range[0.37, 0.41]
![Page 48: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/48.jpg)
2-Matching histograms
?
Z(x,y)
Y(x,y)
We look for a transformation of the image Y Y’ = f (Y) Such that Hist(Y) = Hist(f(Z)) Problem: there are infinitely many functions that can do this transformation. A natural choice is to use f being: - pointwise non linearity - stationary - monotonic (most of the time invertible)
![Page 49: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/49.jpg)
2-Matching histograms
Y’= 0.5 Y= 0.8
Y’ = f (Y)
The function f is just a look up table: it says, change all the pixels of value Y into a value f(Y).
Original intensity
New intensity
Y(x,y)
![Page 50: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/50.jpg)
2-Matching histograms
Y’ = f (Y)
![Page 51: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/51.jpg)
Another example: Matching histograms
10% of pixels are black and 90% are white
5% of pixels have an intensity value within the range[0.37, 0.41]
? ?
![Page 52: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/52.jpg)
Another example: Matching histograms
Y= 0.8
Y’ = f (Y)
The function f is just a look up table: it says, change all the pixels of value Y into a value f(Y).
Original intensity
Y(x,y) Y’= 1 New intensity
![Page 53: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/53.jpg)
Another example: Matching histograms
Y’ = f (Y)
In this example, f is a step function.
![Page 54: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/54.jpg)
Matching histograms of a subband
![Page 55: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/55.jpg)
Y’ = f (Y)
Matching histograms of a subband
![Page 56: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/56.jpg)
Texture analysis
Input texture
(histogram)
Wavelet decomposition (steerable pyr) (histogram)
The texture is represented as a collection of marginal histograms.
(Steerable pyr; Freeman & Adelson, 91)
![Page 57: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/57.jpg)
Texture synthesis
Input texture
(histogram)
(histogram) Heeger and Bergen, 1995
![Page 58: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/58.jpg)
Why does it work? (sort of)
![Page 59: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/59.jpg)
Why does it work? (sort of)
The black and white blocks appear by thresholding (f) a blobby image
Iteration 0
…
Filter bank
![Page 60: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/60.jpg)
Why does it work? (sort of) The black and white blocks appear by thresholding (f) a blobby image
![Page 61: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/61.jpg)
Why does it work? (sort of)
After 6 iterations
Histograms match ok
red = target histogram, blue = current iteration
![Page 62: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/62.jpg)
Color textures
R
G
B
Three textures
![Page 63: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/63.jpg)
Color textures
R
G
B
![Page 64: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/64.jpg)
Color textures
R
G
B
This does not work
![Page 65: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/65.jpg)
Color textures
Problem: we create new colors not present in the original image. Why? Color channels are not independent.
R G B
![Page 66: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/66.jpg)
PCA and decorrelation
R G B
R
G
In the original image, R and G are correlated, but, after synthesis,…
R
G
![Page 67: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/67.jpg)
PCA and decorrelation
R
G
The texture synthesis algorithm assumes that the channels are independent. What we want to do is some rotation
See that in this rotated space, if I specify one coordinate the other remains unconstrained.
U1
U2
Rotation
![Page 68: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/68.jpg)
PCA and decorrelation
R
G
U1
U2
1.0000 0.9303 0.6034
0.9303 0.9438 0.6620
0.6034 0.6620 0.5569
C =
correlation(R,G)
C = D D’ 0.6347 0.6072 0.4779
0.6306 -0.0496 -0.7745
0.4466 -0.7930 0.4144
D =
= 3 x Npixels 3 x Npixels 3 x 3
D’ R G B
U1 U2 U3
PCA finds the principal directions of variation of the data. It gives a decomposition of the covariance matrix as:
By transforming the original data (RGB) using D we get:
The new components (U1,U2,U3) are decorrelated.
![Page 69: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/69.jpg)
Color textures
R
G
B
Rotation Matrix (3x3)
These three textures look similar (high dependency)
These three textures Look less similar (lower dependency)
D’
![Page 70: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/70.jpg)
Color textures
Inverse Rotation Matrix
R
G
B D
![Page 71: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/71.jpg)
Color textures
R
G
B
Rotation Matrix
These three textures look similar (high dependency)
These three textures Look less similar (lower dependency)
D’
Inverse Rotation
R
G
B D
![Page 72: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/72.jpg)
Color channels
Without PCA With PCA
![Page 73: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/73.jpg)
Color channels
![Page 74: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/74.jpg)
Color channels
![Page 75: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/75.jpg)
Examples from the paper
Heeger and Bergen, 1995
![Page 76: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/76.jpg)
Examples from the paper
![Page 77: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/77.jpg)
Examples not from the paper
Input texture
Synthetic texture
But, does it really work even when it seems to work?
![Page 78: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/78.jpg)
But, does it really work??? How to measure how well the representation
constraints the set of equivalent textures?
?
All the textures in this set have the same parameters.
?
? ? ?
![Page 79: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/79.jpg)
How to identify the set of equivalent textures?
?
This does not reveal how poor the representation actually is.
![Page 80: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/80.jpg)
We need a space that is more perceptual In a perceptual space all these noise images are very close. But in pixel space, they are very far away.
How big is this set in a pixels space?
![Page 81: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/81.jpg)
We need a space that is more perceptual In a perceptual space all these noise images are very close. But in pixel space, they are very far away.
How big is this set in a perceptual space?
How big is this set in a pixels space?
![Page 82: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/82.jpg)
These trajectories are more perceptually salient
How to identify the set of equivalent textures?
This set is huge
![Page 83: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/83.jpg)
How to identify the set of equivalent textures?
![Page 84: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/84.jpg)
These trajectories are more perceptually salient
How to identify the set of equivalent textures?
![Page 85: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/85.jpg)
Portilla and Simoncelli
• Parametric representation, based on Gaussian scale mixture prior model for images.
• About 1000 numbers to describe a texture. • Ok results; maybe as good as DeBonet.
![Page 86: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/86.jpg)
Portilla and Simoncelli
![Page 87: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/87.jpg)
Portilla & Simoncelli
Heeger & Bergen Portilla & Simoncelli
![Page 88: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/88.jpg)
Now they look good, but maybe they look too good…
How to identify the set of equivalent textures?
Portilla & Simoncelli
![Page 89: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/89.jpg)
Journal of Vision November 19, 2009 vol. 9 no. 12
![Page 90: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/90.jpg)
Zhu, Wu, & Mumford, 1998
• Principled approach. Based on an assumption of heavy-tailed distributions for an over-complete set of filters.
• Synthesis quality not great, but ok.
![Page 91: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/91.jpg)
Zhu, Wu, & Mumford
• Cheetah Synthetic
![Page 92: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/92.jpg)
De Bonet (and Viola) SIGGRAPH 1997
![Page 93: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/93.jpg)
DeBonet
Learn: use filter conditional statistics across scale.
![Page 94: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/94.jpg)
DeBonet
![Page 95: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/95.jpg)
DeBonet
![Page 96: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/96.jpg)
Two big families of models
1- Parametric models of filter outputs
2- Example-based non-parametric models
![Page 97: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/97.jpg)
![Page 98: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/98.jpg)
Efros & Leung Algorithm
Assuming Markov property, compute P(p|N(p)) – Building explicit probability tables infeasible
p
Synthesizing a pixel
non-parametric sampling
Input image
– Instead, we search the input image for all similar neighborhoods — that’s our pdf for p
– To sample from this pdf, just pick one match at random
![Page 99: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/99.jpg)
Neighborhood Window
input
![Page 100: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/100.jpg)
Varying Window Size
Increasing window size
![Page 101: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/101.jpg)
Synthesis Results french canvas rafia weave
![Page 102: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/102.jpg)
More Results white bread brick wall
![Page 103: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/103.jpg)
Homage to Shannon
![Page 104: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/104.jpg)
Hole Filling
![Page 105: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/105.jpg)
Extrapolation
![Page 106: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/106.jpg)
p
Image Quilting [Efros & Freeman]
• Observation: neighbor pixels are highly correlated
Input image
non-parametric sampling
B
Idea: unit of synthesis = block • Exactly the same but now we want P(B|N(B))
• Much faster: synthesize all pixels in a block at once
• Not the same as multi-scale!
Synthesizing a block
![Page 107: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/107.jpg)
Input texture
B1 B2
Random placement of blocks
block
B1 B2
Neighboring blocks constrained by overlap
B1 B2
Minimal error boundary cut
![Page 108: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/108.jpg)
min. error boundary
Minimal error boundary overlapping blocks vertical boundary
_ = 2
overlap error
![Page 109: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/109.jpg)
Texture Transfer • Take the texture from one
object and “paint” it onto another object – This requires separating
texture and shape – That’s HARD, but we can
cheat – Assume we can capture shape
by boundary and rough shading
• Then, just add another constraint when sampling: similarity to underlying image at that spot
![Page 110: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/110.jpg)
+ =
+ =
parmesan
rice
![Page 111: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/111.jpg)
+ =
![Page 112: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/112.jpg)
= +
![Page 113: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/113.jpg)
Source texture
Target image
Source correspondence
image
Target correspondence image
![Page 114: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/114.jpg)
+ =
![Page 115: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/115.jpg)
Project ideas Non stationary texture synthesis
![Page 116: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/116.jpg)
Project ideas: 3D textures
![Page 117: Lecture 8 - Massachusetts Institute of Technology6.869.csail.mit.edu/sp11/lectures/lecture8.pdf · Lecture 8 . Textures . What is a texture? Which textures are we going to talk about](https://reader033.vdocuments.site/reader033/viewer/2022060406/5f0f622f7e708231d443e3b6/html5/thumbnails/117.jpg)
Project ideas: 3D textures
Can you create a 3D volume that you can navigate? Assume that all slices should have the same statistics. Need knowledge about alpha map?
This is not a solid texture. This is a 3D scene texture.