S. Mandayam/ DIP/ECE Dept.

/Rowan University

Digital Image Processing

ECE.09.452/ECE.09.552

Fall 2009 Lecture 6 October 19, 2009 Shreekanth Mandayam

ECE Department Rowan University

http://engineering.rowan.edu/~shreek/fall09/dip/

S. Mandayam/ DIP/ECE Dept./Rowan University

Plan
Digital Image Restoration
Recall: Environmental Models
Image Degradation Model Image Restoration Model Point Spread Function (PSF) Models

Linear Algebraic Restoration

Unconstrained (Inverse Filter, Pseudoinverse Filter) Constrained (Wiener Filter, Kalman Filter)

Continue Lab 2

S. Mandayam/ DIP/ECE Dept./Rowan University

DIP: Details
Digital Image Processing Digital Image Characteristics Spatial Spectral

Gray-level

Histogram Pre-Processing

DFT

DCT

Enhancement

Restoration

Point Processing

Masking

Filtering

Degradation Models
Compression

Inverse Filtering

Wiener Filtering

Information Theory

Lossless

Lossy

LZW (gif)
Segmentation

Transform-based (jpeg)

Edge Detection

Description

Shape Descriptors

Texture

Morphology

S. Mandayam/ DIP/ECE Dept./Rowan University

Image Preprocessing
Enhancement Restoration
Inverse filtering Wiener filtering

Spatial Domain

Spectral Domain Filtering >>fft2/ifft2 >>fftshift

Point Processing >>imadjust >>histeq

Spatial filtering >>filter2

S. Mandayam/ DIP/ECE Dept./Rowan University

Enhancement vs. Restoration

Better visual representation
Subjective No quantitative measures

Remove effects of sensing environment

Objective Mathematical, model dependent quantitative measures

S. Mandayam/ DIP/ECE Dept./Rowan University

Degradation Model
f(x,y) h(x,y) S n(x,y) Degradation Model: g = h*f + n g(x,y)

demos/demo5blur_invfilter/ demos/demo5blur_invfilter/degrade.m

S. Mandayam/ DIP/ECE Dept./Rowan University

Restoration Model
f(x,y)

Degradation Model

Restoration Filter

f(x,y)

Unconstrained
Inverse Filter Pseudo-inverse Filter

Constrained
Wiener Filter

demos/demo5blur_invfilter/

S. Mandayam/ DIP/ECE Dept./Rowan University

f(x,y) Build degradation model Analyze using algebraic techniques Formulate restoration algorithms Implement using Fourier transforms f(x,y)

Approach
g = h*f + n

g = Hf + n -1 -1 -1 W g = DW f + W n f = H -1 g

F(u,v) = G(u,v)/H(u,v)
demos/demo5blur_invfilter/

S. Mandayam/ DIP/ECE Dept./Rowan University

Degradation & Restoration Examples: Gonzalez & Woods

Atmospheric Turbulence Model

S. Mandayam/ DIP/ECE Dept./Rowan University

Degradation & Restoration Examples: Gonzalez & Woods

Example 5.11: Inverse Filtering

S. Mandayam/ DIP/ECE Dept./Rowan University

Degradation & Restoration Examples: Gonzalez & Woods

Example 5.12: Wiener Filtering

S. Mandayam/ DIP/ECE Dept./Rowan University

Degradation & Restoration Examples: Gonzalez & Woods

Example 5.10: Planar Motion Model

S. Mandayam/ DIP/ECE Dept./Rowan University

Degradation & Restoration Examples: Gonzalez & Woods

Example 5.13: Inverse and Wiener Filtering

S. Mandayam/ DIP/ECE Dept./Rowan University

Lab 3: Degradation Models and Digital Image Restoration

http://engineering.rowan.edu/~shreek/fall09/dip/lab3.html

S. Mandayam/ DIP/ECE Dept./Rowan University

Class on Oct 26
At the South Jersey Tech Park, www.sjtechpark.org, (On Rt 322-W at the Rt. 55 interchange) Guest Lecture by George Lecakes: Modeling and Visualization for Virtual Reality Includes demo of the CAVE

S. Mandayam/ DIP/ECE Dept./Rowan University

Class on Nov 2
Continue Lab Project 3 http://engineering.rowan.edu/~shreek /fall09/dip/lab3.html

S. Mandayam/ DIP/ECE Dept./Rowan University

Summary