What are Good Apertures for Defocus Deblurring? - PowerPoint PPT Presentation

1 / 34
About This Presentation
Title:

What are Good Apertures for Defocus Deblurring?

Description:

For white noise and L2 distance, Weiner filter is optimal ... Annular. Image(binary) Image(gray) Random. Outline - Evaluation Criterion for Aperture Patterns ... – PowerPoint PPT presentation

Number of Views:33
Avg rating:3.0/5.0
Slides: 35
Provided by: Chan93
Category:

less

Transcript and Presenter's Notes

Title: What are Good Apertures for Defocus Deblurring?


1
What are Good Apertures for Defocus Deblurring?
Changyin Zhou Shree K. Nayar
Columbia University ICCP 2009, San Francisco
2
Geometry of Defocus
Lens
Object
Sensor
3
Formulation of Defocus
In the Spatial Domain
4
Formulation of Defocus
In the Fourier Domain
Focused Image
Captured Image
PSF
Image Noise
5
Various Aperture Coding Techniques
6
What are Good Apertures for Defocus Deblurring?
7
Outline
- Background
- Evaluation Criterion for Aperture Patterns
- Pattern Optimization
- Experiments
8
How to Evaluate Aperture Patterns?
Captured Image
Focused Image
9
How to Evaluate Aperture Patterns?
Captured Image
Focused Image
10
The Optimal Linear Deblurring Algorithm
  • For white noise and L2 distance, Weiner filter
    is optimal

11
The Optimal Linear Deblurring Algorithm

12
How to Evaluate Aperture Patterns?
Captured Image
Focused Image
An Optimal Deblurring Algorithm
Deblurred Image
Linear system
13
Evaluation Criterion for Aperture Pattern
14
Evaluate Patterns Using the Criterion
R Curve of Circular Pattern
R
Noise Level
Noise Level
15
Evaluate Patterns Using the Criterion
Curves of Relative R R(K)/R(K0)
Noise Level
  • Coded apertures help more when noise level is
    low.
  • Circular aperture is better when noise level is
    high.

16
Outline
- Background
- Evaluation Criterion for Aperture Patterns
- Pattern Optimization
- Experiments
17
Pattern Optimization
  • Difficult to solve analytically
  • Patterns evaluated in the Fourier domain, but
    strictly constrained in the spatial domain
    .
  • Difficult to do brute force search
  • For binary patterns of resolution N x N, the
    number of possible solutions is huge,

when N 13, if evaluating one pattern takes 1
millisecond, the brute force search requires
1045 yrs.
18
Pattern Optimization
Pattern Evolution in Genetic Algorithm 13 x 13
binary patterns 8 different noise levels
2nd Run
3nd Run
19
Evaluate the Optimized Patterns
20
Evaluate the Optimized Patterns
Relative R curves of the Optimized Patterns
Image
s 0.001
21
Outline
- Background
- Evaluation Criterion for Aperture Patterns
- Pattern Optimization
- Experiments
22
Implementation
Precision Laser Photoplot (1 micron)
23
Implementation
Canon EF 50mm f/1.8 Lens
24
Implementation
Image Pattern
Veeraraghavan et al.s Pattern
Levin et al.s Pattern
Circular Pattern (wide open)
Our Optimized Pattern
25
Comparison Experiments on a CZP Chart
Captured Images
Focused Image
Circular Pattern
Veeraraghavan et als Pattern
Image Pattern
26
Comparison Experiments on a CZP Chart
Deblurred Images
Focused Image
Circular Pattern
Levin et al.s Pattern
Veeraraghavan et als Pattern
Image Pattern
27
Shrek
Captured Image using the Optimized Pattern
28
Shrek
Captured Image using the Optimized Pattern
Deblurring Result
29
On the Street
Captured Image Using the Optimized Pattern
30
On the Street
Captured Image Using the Optimized Pattern
Deblurring Result
31
Traffic Scene
Captured Image Using the Optimized Pattern
32
Traffic Scene
Captured Image Using the Optimized Pattern
Deblurring Result
33
Summary
  • Main Contributions
  • Aperture Evaluation Criterion for Defocus
    Deblurring
  • Aperture Pattern Optimization for Defocus
    Deblurring
  • Future Work
  • Optimize gray-level patterns
  • Apply Criterion to other PSF Engineering Problems
  • Extend Analysis to Account for Diffraction

34
Thank You!
35
Genetic Algorithm for Pattern Optimization
36
Genetic Algorithm for Pattern Optimization
Noise level 0.005
1st Run
2nd Run
3nd Run
Write a Comment
User Comments (0)
About PowerShow.com