Fast and High Quality Overlap Repair for Patch-Based Texture Synthesis Andrew Nealen Marc Alexa Discrete Geometric Modeling Group (DGM) Technische Universität Darmstadt Andrew Nealen and Marc.

Download Report

Transcript Fast and High Quality Overlap Repair for Patch-Based Texture Synthesis Andrew Nealen Marc Alexa Discrete Geometric Modeling Group (DGM) Technische Universität Darmstadt Andrew Nealen and Marc.

Fast and High Quality Overlap Repair for Patch-Based Texture Synthesis

Andrew Nealen Marc Alexa

Discrete Geometric Modeling Group (DGM) Technische Universität Darmstadt Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Our Setting: 2D Texture Synthesis

n x m Input Texture N x M Output Texture

The goal: Synthesize an output texture which is

perceptually similar

to the input texture. Also ensure that the result contains

sufficient variation

.

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Patch-Based Texture Synthesis

Some Existing Methods

A Very Popular 2D Texture Synthesis Method

• Image Quilting [Efros and Freeman 2001] • Graphcut Texures [Kwatra et. al 2003] • Wang Tiles [Cohen et. al 2003] Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Patch-Based Texture Synthesis

Some Existing Methods

A Very Popular 2D Texture Synthesis Method

Image Quilting [Efros and Freeman 2001]

• Graphcut Texures [Kwatra et. al 2003] • Wang Tiles [Cohen et. al 2003] Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Patch-Based Texture Synthesis

Some Existing Methods

A Very Popular 2D Texture Synthesis Method

Image Quilting [Efros and Freeman 2001]

• Graphcut Texures [Kwatra et. al 2003] • Wang Tiles [Cohen et. al 2003]

A B

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Patch-Based Texture Synthesis

Some Existing Methods

A Very Popular 2D Texture Synthesis Method

Image Quilting [Efros and Freeman 2001]

• Graphcut Texures [Kwatra et. al 2003] • Wang Tiles [Cohen et. al 2003]

A B

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Patch-Based Texture Synthesis

Some Existing Methods

A Very Popular 2D Texture Synthesis Method

Image Quilting [Efros and Freeman 2001]

• Graphcut Texures [Kwatra et. al 2003] • Wang Tiles [Cohen et. al 2003]

A B

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Patch-Based Texture Synthesis

Some Existing Methods

A Very Popular 2D Texture Synthesis Method

Image Quilting [Efros and Freeman 2001]

• Graphcut Texures [Kwatra et. al 2003] • Wang Tiles [Cohen et. al 2003]

A B

_ 2 Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Patch-Based Texture Synthesis

Some Existing Methods

A Very Popular 2D Texture Synthesis Method

Image Quilting [Efros and Freeman 2001]

• Graphcut Texures [Kwatra et. al 2003] • Wang Tiles [Cohen et. al 2003]

A B

_ 2

= overlap error

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Patch-Based Texture Synthesis

Some Existing Methods

A Very Popular 2D Texture Synthesis Method

Image Quilting [Efros and Freeman 2001]

• Graphcut Texures [Kwatra et. al 2003] • Wang Tiles [Cohen et. al 2003]

A B

_ 2

= overlap error

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Patch-Based Texture Synthesis

Some Existing Methods

A Very Popular 2D Texture Synthesis Method

Image Quilting [Efros and Freeman 2001]

• Graphcut Texures [Kwatra et. al 2003] • Wang Tiles [Cohen et. al 2003]

A B

_ 2

= overlap error

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Patch-Based Texture Synthesis

Some Existing Methods

A Very Popular 2D Texture Synthesis Method

Image Quilting [Efros and Freeman 2001]

• Graphcut Texures [Kwatra et. al 2003] • Wang Tiles [Cohen et. al 2003]

A B

_ 2

= overlap error

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Patch-Based Texture Synthesis

Some Existing Methods

A Very Popular 2D Texture Synthesis Method

• Image Quilting [Efros and Freeman 2001] •

Graphcut Texures [Kwatra et. al 2003]

• Wang Tiles [Cohen et. al 2003] Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Patch-Based Texture Synthesis

Some Existing Methods

A Very Popular 2D Texture Synthesis Method

• Image Quilting [Efros and Freeman 2001] • Graphcut Texures [Kwatra et. al 2003] •

Wang Tiles [Cohen et. al 2003]

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Patch-Based Texture Synthesis

Hybrid Texture Synthesis (HTS)

Introduced at EGSR 2003 [Nealen and Alexa]

Adaptive Patch Sampling, like Mapping [Soler et. al 2002] Hierarchical Pattern • Per-Pixel Overlap Re-synthesis Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Hybrid Texture Synthesis

Method Input (n x m) Intermediate Result Goal: From nxm, synthesize NxM similar, but not identical

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Hybrid Texture Synthesis

Method Input (n x m) Intermediate Result Goal: From nxm, synthesize NxM similar, but not identical

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Hybrid Texture Synthesis

Method Patch-Search in the Input + Copy to Result + Mark Invalid Pixels Input (n x m) Intermediate Result Goal: From nxm, synthesize NxM similar, but not identical

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Hybrid Texture Synthesis

Method Patch-Search in the Input + Copy to Result + Mark Invalid Pixels Input (n x m) Intermediate Result Goal: From nxm, synthesize NxM similar, but not identical Per-Pixel Re-synthesis Steps (for each Patch)

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Hybrid Texture Synthesis

Method Patch-Search in the Input + Copy to Result + Mark Invalid Pixels Input (n x m) Intermediate Result Goal: From nxm, synthesize NxM similar, but not identical Per-Pixel Re-synthesis Steps (for each Patch)

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Hybrid Texture Synthesis

Method Patch-Search in the Input + Copy to Result + Mark Invalid Pixels Input (n x m) Intermediate Result Goal: From nxm, synthesize NxM similar, but not identical Per-Pixel Re-synthesis Steps (for each Patch)

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Hybrid Texture Synthesis

Method Patch-Search in the Input + Copy to Result + Mark Invalid Pixels Input (n x m) Intermediate Result Goal: From nxm, synthesize NxM similar, but not identical Per-Pixel Re-synthesis Steps (for each Patch)

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Hybrid Texture Synthesis

Method Patch-Search in the Input + Copy to Result + Mark Invalid Pixels Input (n x m) Intermediate Result Goal: From nxm, synthesize NxM similar, but not identical Per-Pixel Re-synthesis Steps (for each Patch)

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Hybrid Texture Synthesis

Generalization: Pro and Con

Pro: General Method for Overlap Repair

• Complementary to other Methods, such as Minimum Error-Boundary-Cut (MEBC) or Feathering, yet oftentimes produces better results • Generalizes to arbitrary patch shapes, i.e. is applicable to Graphcut Textures, Wang Tiles, etc.

Con: Computationally Expensive

• Exhaustive search for each based on mostly irregular invalid valid pixel in the overlap, neighborhood • Has O(rN log N) complexity -> Doesn‘t scale well.

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Hybrid Texture Synthesis

Generalization: Pro and Con

Pro: General Method for Overlap Repair

• Complementary to other Methods, such as Minimum Error-Boundary-Cut (MEBC) or Feathering, yet oftentimes produces better results • Generalizes to arbitrary patch shapes, i.e. is applicable to Graphcut Textures, Wang Tiles, etc.

Con: Computationally Expensive

• Exhaustive search for each based on mostly irregular invalid valid pixel in the overlap, neighborhood • Has O(rN log N) complexity -> Doesn‘t scale well.

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Hybrid Texture Synthesis

Generalization: Pro and Con

Pro: General Method for Overlap Repair

• Complementary to other Methods, such as Minimum Error-Boundary-Cut (MEBC) or Feathering, yet oftentimes produces better results • Generalizes to arbitrary patch shapes, i.e. is applicable to Graphcut Textures, Wang Tiles, etc.

Con: Computationally Expensive

• Exhaustive search for each based on mostly irregular invalid valid pixel in the overlap, neighborhood • Has O(rN log N) complexity -> Doesn‘t scale well.

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Fast Overlap Repair

Basic Idea

Inspiration

• • Ashikhmin: Synthesizing Natural Textures [2001] termed Coherence Search Tong et. al‘s extension: k-Coherence Search [2002]

Basic Idea: Intelligently Reduce Search Space

• Only search within a set of coherent pixels • Introduce Trade-off between quality and speed Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Fast Overlap Repair

Coherence Search

Applying Coherence Search

• For each pixel in the output, store its location in the input in a source map (same size as the output texture)

Input Texture Intermediate Result + Source Map

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Fast Overlap Repair

Coherence Search

Applying Coherence Search

• When searching for a new pixel, only consider input pixels which are coherent with neighboring output pixels

Source Map Lookup Input Texture Intermediate Result + Source Map

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Fast Overlap Repair

Coherence Search

Applying Coherence Search

• When searching for a new pixel, only consider input pixels which are coherent with neighboring output pixels

Source Map Lookup Input Texture Intermediate Result + Source Map Consequence in this example: Only two possible candidates

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Fast Overlap Repair

Coherence Search

Applying Coherence Search

• Simply comparing to the coherent pixels results in seams similar to Image Quilting (MEBC)

Example: 64x64 Texture Synthesized from four 32x32 Patches Coherence

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Exhaustive

Fast Overlap Repair

Coherence Search

Applying Coherence Search

• Simply comparing to the coherent pixels results in seams similar to Image Quilting (MEBC)

Example: 64x64 Texture Synthesized from four 32x32 Patches Coherence

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Exhaustive

Fast Overlap Repair

Coherence Search

Applying Coherence Search

• Simply comparing to the coherent pixels results in seams similar to Image Quilting (MEBC)

Example: 64x64 Texture Synthesized from four 32x32 Patches Coherence

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Exhaustive

Fast Overlap Repair

k-Coherence Search

Better: Applying k-Coherence Search Source Map Lookup Input Texture Intermediate Result + Source Map

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Fast Overlap Repair

k-Coherence Search

Better: Applying k-Coherence Search

• Extend the set by the k-nearest neighbors (knn) of each coherent pixel (in feature space) and remove duplicates

Source Map Lookup Input Texture Intermediate Result + Source Map

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Fast Overlap Repair

k-Coherence Search

Precomputation of knn Data Structure

• • • Performed once for each nxm input texture and stored for repeated use User defines size of box-shaped neighborhood n

p

For each of the nxm input pixels • • ─ Construct feature vector by ordered concatenation of the

n p

x

n p

RGB-triples in the box-shaped neighborhood Dimension reduction (75-90%) by applying PCA Compute indices of k-nearest neighbors to each pixel Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Fast Overlap Repair

k-Coherence Search

Source Map Maintenance

• Each valid pixel in the overlap region is a linear blend (feathering) of at least two original pixel values, i.e. from at least two different sources • To avoid the maintenance of multiple source maps, simply store the source of the pixel with greatest contribution in a single

source map

Blue: invalid overlap pixels Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Results

varying k k = 4 k = 1 k = 11

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Exhaustive

Results

varying k k = 4 k = 1 k = 11

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Exhaustive

Results

varying k k = 4 k = 1 k = 11

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Exhaustive

Results

varying k k = 4 k = 1 k = 11

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Exhaustive

Results

timings Input Exhaustive n = 7x7 k-Coherence n = 3x3 | k = 5 k-Coherence n = 5x5 | k = 11 scales 64 × 64 δ max Δ max = 0.02 = 0.05

Pre: 0 sec.

Synth: 283 sec.

Pre: 6+3 sec.

Synth: 226 sec.

Pre: 6+4 sec.

Synth: 427 sec.

rock 128 × 128 δ max Δ max = 0.02 = 0.05

Pre: 0 sec.

Synth: 533 sec.

stonewall 200 × 200 δ max Δ max = 0.02 = 0.03

Pre: 0 sec.

Synth: 985 sec.

Pre: 45+62 sec.

Synth: 226 sec.

Pre: 45+74 sec.

Synth: 415 sec.

Pre: 247+28 s Synth: 178 sec.

Pre: 247+37 s Synth: 350 sec.

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Results

timings Input Exhaustive n = 7x7 k-Coherence n = 3x3 | k = 5 k-Coherence n = 5x5 | k = 11 scales 64 × 64 δ max Δ max = 0.02 = 0.05

Pre: 0 sec.

Synth: 283 sec.

Pre: 6+3 sec.

Synth: 226 sec.

Pre: 6+4 sec.

Synth: 427 sec.

rock 128 × 128 δ max Δ max = 0.02 = 0.05

Pre: 0 sec.

Synth: 533 sec.

stonewall 200 × 200 δ max Δ max = 0.02 = 0.03

Pre: 0 sec.

Synth: 985 sec.

Pre: 45+62 sec.

Synth: 226 sec.

Pre: 45+74 sec.

Synth: 415 sec.

Pre: 247+28 s Synth: 178 sec.

Pre: 247+37 s Synth: 350 sec.

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Results

timings Input Exhaustive n = 7x7 k-Coherence n = 3x3 | k = 5 k-Coherence n = 5x5 | k = 11 scales 64 × 64 δ max Δ max = 0.02 = 0.05

Pre: 0 sec.

Synth: 283 sec.

Pre: 6+3 sec.

Synth: 226 sec.

Pre: 6+4 sec.

Synth: 427 sec.

rock 128 × 128 δ max Δ max = 0.02 = 0.05

Pre: 0 sec.

Synth: 533 sec.

stonewall 200 × 200 δ max Δ max = 0.02 = 0.03

Pre: 0 sec.

Synth: 985 sec.

Pre: 45+62 sec.

Synth: 226 sec.

Pre: 45+74 sec.

Synth: 415 sec.

Pre: 247+28 s Synth: 178 sec.

Pre: 247+37 s Synth: 350 sec.

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Results

timings Input Exhaustive n = 7x7 k-Coherence n = 3x3 | k = 5 k-Coherence n = 5x5 | k = 11 scales 64 × 64 δ max Δ max = 0.02 = 0.05

Pre: 0 sec.

Synth: 283 sec.

Pre: 6+3 sec.

Synth: 226 sec.

Pre: 6+4 sec.

Synth: 427 sec.

rock 128 × 128 δ max Δ max = 0.02 = 0.05

Pre: 0 sec.

Synth: 533 sec.

stonewall 200 × 200 δ max Δ max = 0.02 = 0.03

Pre: 0 sec.

Synth: 985 sec.

Pre: 45+62 sec.

Synth: 226 sec.

Pre: 45+74 sec.

Synth: 415 sec.

Pre: 247+28 s Synth: 178 sec.

Pre: 247+37 s Synth: 350 sec.

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Results

timings Input Exhaustive n = 7x7 k-Coherence n = 3x3 | k = 5 k-Coherence n = 5x5 | k = 11 scales 64 × 64 δ max Δ max = 0.02 = 0.05

Pre: 0 sec.

Synth: 283 sec.

Pre: 6+3 sec.

Synth: 226 sec.

Pre: 6+4 sec.

Synth: 427 sec.

rock 128 × 128 δ max Δ max = 0.02 = 0.05

Pre: 0 sec.

Synth: 533 sec.

stonewall 200 × 200 δ max Δ max = 0.02 = 0.03

Pre: 0 sec.

Synth: 985 sec.

Pre: 45+62 sec.

Synth: 226 sec.

Pre: 45+74 sec.

Synth: 415 sec.

Pre: 247+28 s Synth: 178 sec.

Pre: 247+37 s Synth: 350 sec.

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Results

timings Input Exhaustive n = 7x7 k-Coherence n = 3x3 | k = 5 k-Coherence n = 5x5 | k = 11 scales 64 × 64 δ max Δ max = 0.02 = 0.05

Pre: 0 sec.

Synth: 283 sec.

Pre: 6+3 sec.

Synth: 226 sec.

Pre: 6+4 sec.

Synth: 427 sec.

rock 128 × 128 δ max Δ max = 0.02 = 0.05

Pre: 0 sec.

Synth: 533 sec.

stonewall 200 × 200 δ max Δ max = 0.02 = 0.03

Pre: 0 sec.

Synth: 985 sec.

Pre: 45+62 sec.

Synth: 226 sec.

Pre: 45+74 sec.

Synth: 415 sec.

Pre: 247+28 s Synth: 178 sec.

Pre: 247+37 s Synth: 350 sec.

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Results

timings Input Exhaustive n = 7x7 k-Coherence n = 3x3 | k = 5 k-Coherence n = 5x5 | k = 11 scales 64 × 64 δ max Δ max = 0.02 = 0.05

Pre: 0 sec.

Synth: 283 sec.

Pre: 6+3 sec.

Synth: 226 sec.

Pre: 6+4 sec.

Synth: 427 sec.

rock 128 × 128 δ max Δ max = 0.02 = 0.05

Pre: 0 sec.

Synth: 533 sec.

stonewall 200 × 200 δ max Δ max = 0.02 = 0.03

Pre: 0 sec.

Synth: 985 sec.

Pre: 45+62 sec.

Synth: 226 sec.

Pre: 45+74 sec.

Synth: 415 sec.

Pre: 247+28 s Synth: 178 sec.

Pre: 247+37 s Synth: 350 sec.

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Results

timings Input Exhaustive n = 7x7 k-Coherence n = 3x3 | k = 5 k-Coherence n = 5x5 | k = 11 scales 64 × 64 δ max Δ max = 0.02 = 0.05

Pre: 0 sec.

Synth: 283 sec.

Pre: 6+3 sec.

Synth: 226 sec.

Pre: 6+4 sec.

Synth: 427 sec.

rock 128 × 128 δ max Δ max = 0.02 = 0.05

Pre: 0 sec.

Synth: 533 sec.

stonewall 200 × 200 δ max Δ max = 0.02 = 0.03

Pre: 0 sec.

Synth: 985 sec.

Pre: 45+62 sec.

Synth: 226 sec.

Pre: 45+74 sec.

Synth: 415 sec.

Pre: 247+28 s Synth: 178 sec.

Pre: 247+37 s Synth: 350 sec.

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Results

timings Input Exhaustive n = 7x7 k-Coherence n = 3x3 | k = 5 k-Coherence n = 5x5 | k = 11 scales 64 × 64 δ max Δ max = 0.02 = 0.05

Pre: 0 sec.

Synth: 283 sec.

Pre: 6+3 sec.

Synth: 226 sec.

Pre: 6+4 sec.

Synth: 427 sec.

rock 128 × 128 δ max Δ max = 0.02 = 0.05

Pre: 0 sec.

Synth: 533 sec.

stonewall 200 × 200 δ max Δ max = 0.02 = 0.03

Pre: 0 sec.

Synth: 985 sec.

Pre: 45+62 sec.

Synth: 226 sec.

Pre: 45+74 sec.

Synth: 415 sec.

Pre: 247+28 s Synth: 178 sec.

Pre: 247+37 s Synth: 350 sec.

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Results

timings Input Exhaustive n = 7x7 k-Coherence n = 3x3 | k = 5 k-Coherence n = 5x5 | k = 11 scales 64 × 64 δ max Δ max = 0.02 = 0.05

Pre: 0 sec.

Synth: 283 sec.

Pre: 6+3 sec.

Synth: 226 sec.

Pre: 6+4 sec.

Synth: 427 sec.

rock 128 × 128 δ max Δ max = 0.02 = 0.05

Pre: 0 sec.

Synth: 533 sec.

stonewall 200 × 200 δ max Δ max = 0.02 = 0.03

Pre: 0 sec.

Synth: 985 sec.

Pre: 45+62 sec.

Synth: 226 sec.

Pre: 45+74 sec.

Synth: 415 sec.

Pre: 247+28 s Synth: 178 sec.

Pre: 247+37 s Synth: 350 sec.

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Results

Exhaustive n = 7x7 k-Coherence n = 3x3 | k = 5 k-Coherence n = 5x5 | k = 11

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Results

Exhaustive n = 7x7 k-Coherence n = 3x3 | k = 5 k-Coherence n = 5x5 | k = 11

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Results

Synthesis Comparisons Input Efros/Leung Wei/Levoy IQ PBS

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

HTS

Results

Synthesis Comparisons Input Efros/Leung Wei/Levoy IQ PBS

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

HTS

Results

Synthesis Comparisons Input Efros/Leung Wei/Levoy IQ PBS

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

HTS

Results

Synthesis Comparisons Input Efros/Leung Wei/Levoy IQ PBS

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

HTS

Results

Synthesis Comparisons Input Efros/Leung Wei/Levoy IQ PBS

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

HTS

Results

Synthesis Comparisons Input Efros/Leung Wei/Levoy IQ PBS

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

HTS

Results

Synthesis Comparisons Input Efros/Leung Wei/Levoy IQ PBS

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

HTS

Conclusions and Future Work

Improve Error Metric

• • Still using the L 2 norm due to its simplicity Develop a metric which takes feature mismatch into account • • Texton map approach [Zhang et al. 2003] Feature Map [Wu and Yu 2004] performs even better, and for near-regular textures, see [Liu et. al 2004] (both to appear at

SIGGRAPH

2004) Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004

Questions ?

Contact Information

Andrew Nealen [email protected]

Marc Alexa [email protected]

http://www.dgm.informatik.tu-darmstadt.de

Matlab code: http://www.dgm.informatik.tu-darmstadt.de/research/texsynth.html

Andrew Nealen and Marc Alexa, Discrete Geometric Modeling Group, TU Darmstadt, 2004