dd

Image Processing EECS 556 - Winter 2010

Instructor: Prof. Silvio Savarese
Office hours: by appointment
Webpage: http://www.eecs.umich.edu/~silvio/

Classroom: 1017 DOW
Time: T Th 1:30pm-3:00pm

 

-------------------------------------------------------------------------------------------------------------------

EECS 556- Tentative Schedule (subject to changes)

 

 
Lect. Date Topic Link Details
1 Tues 1.12 Introduction Slides

Course Policies and Introduction to Image Processing

2 Thur 1.14 2D Continous-Space Signals & Systems
Basic concepts and properties

Notes

Slides

signal transformations
signal classes (symmetry, periodic)
2D impulse function / properties
2D systems, input-output relationship
system classifications
point spread function (PSF); impulse response and representation
superposition property
convolution & convolution properties
3 Tues 1.19 2D Continous-Space Signals & Systems
Systems and Fouries Transforms

Notes

Slides

system properties in terms of PSF
orthogonal representation of signals
2D Fourier series
2D Fourier transform: Properties of 2D FT ;New properties
OTF/MTF
2D-FT Examples
rms bandwidth, rms time duration
time-bandwidth product, gaussian example
4 Thur 1.21

Sampling

2D Discrete-Space Signals & Systems; Basic concepts and properties I

Notes

Notes

Slides

rectilinear sampling; nyquist sampling rate; sampling theorem; aliasing
sinc interpolation;
video interpolation
2D-DSFT / 2D-FT relationship; 2D DFT/FFT, relationships to 2D-FT
fftshift

2D Kronecker impulse function / properties
notation, coordinate systems;

5 Tues 1.26

2D Discrete-Space Signals & Systems; Basic concepts and properties I

 



Notes

Slides

Lim Ch. 1

signal classes (symmetry, periodic, separable); circular symmetry
2D systems, input-output relationship
system classifications;
impulse response
convolution sum; graphical convolution; separable convolution; edge effects; convolution properties
correlation; system properties in terms of PSF: causality, memory, stability, invertibility, rotation invariance; eigenfunctions

6 Thur 1.28

2D Discrete-Space Signals & Systems; Basic concepts and properties II

Optical imaging basics

Notes
Lim Ch. 1

Notes

Slides

2D discrete-space Fourier transform; existence/convergence of FT; periodic signals; properties
LSI systems in frequency domain
magnitude vs phase; phase-only reconstruction / magnitude retrieval; sampling revisited
filters; introduction to filter design; separability vs rotation invariance.

Fourier transforming property of lenses
Imaging property of lenses (PSF, frequency response)

7 Tue 2.2 FIR filters

Discrete Fourier Transform; DF series

Notes
Lim Ch. 4

Notes

Slides

FIR filters
zero phase; symmetries; computation
filter specs; ideal filter impulse response; filtering using FFT

Discrete-space orthogonal representation; properties; Discrete Fourier series; discrete-space orthogonal representation properties;
circular convolution; circular convolution example

8 Thur 2.4 Discrete Fourier Transform (DFT)

Notes

Slides

DFT; 3 imperfect perspectives: orthonormal basis, periodization, sampling DSFT; derivation; properties; circular convolution; relationship to DSFT; zero padding; overlap-add method; sampling dsft;
sampling H(omega) example; matrix representation of DFT DCT: motivation; derivation; properties
FFT: row column decomposition

9 Tues 2.9 Image enhancement (I)

Notes

Corrections

Slides

Contrast adjustment; piecewise linear contrast adjustment; Noise smoothing; linear vs median filters; Edge detection; gradient-based methods; derivatives from discrete images; laplacian-based methods;

10 Thur 2.11 Image enhancement (II)

Notes

Slides

parametric edge-detection methods; canny's method; Harris corner detector;

11 Tues 2.16 Image enhancement (III)

Notes

Slides

Blob detectors and SIFT/DOG detector; statistical approach; histograms;histogram normalization;
12 Thur 2.18 Motion estimation

Notes

Slides

Motion estimation; Motion-compensated interpolation; region matching methods; space-time constraint equation; normal equation and optical flow
13 Tues 2.23

Interpolation

Notes

Notes on Splines

Slides

Image interpolation; nearest, sinc, bilinear
interpolator properties; b-splines; B-splines; B-spline interpolation; cardinal B-splines; temporal interpolation

  Tues 2.23 Mid Term Exam  

48 hours take home exam - see eecs556.html for details;
Due on Thursday Feb 25 at 7:30pm by prof Savarese's office

14 Thur 2.25 Spectral estimation (I)

Notes

Corrections

Slides

Random vectors; random processes, WSS; autocorrelation, cross-correlation; power spectral density;RP's through LSI systems;
- Tues 3.2 Study Spring Break
- Thur 3.4 Study Spring Break
15 Tues 3.9 Spectral estimation (II)

Notes

Slides

Noncausal Wiener filter; Spectral estimation; periodogram; cross-correlation and windowing; nonnegativity of power spectra;autocorrelation functions for filtered point process
16 Thur 3.11 Introduction to Image restoration (I)

Notes

Slides

Overview; type of noise; noise estimation; degradation estimation; motion blur; Frequency domain methods;
blur removal; inverse form; Wiener filter form; examples

17 Tues 3.16 No Class - ECCV Deadline    
18 Thur 3.18

Introduction to Image restoration (II)

Notes

Slides

Pixel Domain Methods; Matrix Formulation; Least Square solution; SVD; Regularization

19 Tues 3.23

Introduction to Image restoration (III)

Statistical methods for image restoration (I)

Notes

Slides

Regularization; Tikhonov matrix; Space‐Varying Approaches; Adaptive wiener Filtering; Steepest Descent Algorithm;

Statistical methods for image restoration; ML estimation; MAP estimation

20 Thur 3.25 Statistical methods for image restoration (II)

Notes

Slides

Statistical methods for image restoration; Penalized-likelihood estimation; Non-quadratic regularization; Edge preserving methods; Huber Potentials; Examples;
21 Tues 3.30 Statistical methods for image restoration (III)

Notes

Slides

Statistical methods for image restoration; Circulant Analysis; Circulant analysis for the QPWLS;
22 Thur 4.1 Statistical methods for image restoration (IV)

Notes

Notes

Slides

Statistical methods for image restoration; Adaptative Wiener Filters; Other formulations; blind deconvolution;
23 Tues 4.6 Image coding (I)

Notes

Slides

Introduction to image coding; Basic scheme; Scalar and Vector quantization; Codeword assignments
24 Thur 4.8 Image segmentation Slides Image segmentation; Kmean clustering; Meanshift; normalized cuts; Other methods.
25 Tues 4.13

Image coding (II)

Notes

Notes

Slides

Predictive Coding; Transform image coding; Karhunen-Loeve Transform; DCT transform; JPG compression;

 

  Thur 4.15 Project Presentations 1:30-3pm    
  Tues 4.20 Project Presentations 1:30-3pm    
  Wed 4.21 Project Presentations 7:00-8:45pm    
    Project final report due: Apr 28 Midnight