EE123 Digital Signal Processing

Similar documents
Image Reconstruction from Projection

Central Slice Theorem

Modern CT system generations Measurement of attenuation

Reconstruction in CT and hints to the assignments

Digital Image Processing

Biomedical Image Analysis. Spatial Filtering

9. Computed Tomography (CT) 9.1 Absorption of X-ray

Radon Transform and Filtered Backprojection

Continuous Space Fourier Transform (CSFT)

Lecture 5: Frequency Domain Transformations

Lecture 2: 2D Fourier transforms and applications

EECS490: Digital Image Processing. Lecture #16

Fourier analysis and sampling theory

Computer Vision and Graphics (ee2031) Digital Image Processing I

Sampling: Application to 2D Transformations

Digital Image Processing. Lecture 6

BME I5000: Biomedical Imaging

Computer Vision 2. SS 18 Dr. Benjamin Guthier Professur für Bildverarbeitung. Computer Vision 2 Dr. Benjamin Guthier

Digital Signal Processing Lecture Notes 22 November 2010

Tomographic Reconstruction

Topics. Projections. Review Filtered Backprojection Fan Beam Spiral CT Applications. Bioengineering 280A Principles of Biomedical Imaging

Topics. Review Filtered Backprojection Fan Beam Spiral CT Applications. Bioengineering 280A Principles of Biomedical Imaging

Chapter6 Image Reconstruction

GE s Revolution CT MATLAB III: CT. Kathleen Chen March 20, 2018

The main goal of Computer Graphics is to generate 2D images 2D images are continuous 2D functions (or signals)

Joint ICTP-TWAS Workshop on Portable X-ray Analytical Instruments for Cultural Heritage. 29 April - 3 May, 2013

ABSTRACT INTRODUCTION


MEDICAL IMAGING 2nd Part Computed Tomography

Motivation. The main goal of Computer Graphics is to generate 2D images. 2D images are continuous 2D functions (or signals)

Image preprocessing in spatial domain

Computed Tomography (Part 2)

Image preprocessing in spatial domain

LOWELL WEEKLY JOURNAL.

Digital Image Processing

Introduction to Medical Imaging. Cone-Beam CT. Klaus Mueller. Computer Science Department Stony Brook University

Lab 1: Basic operations on images in Matlab

Topics. 2D Signal. a b. c d. Bioengineering 280A Principles of Biomedical Imaging. Fall Quarter 2004 Lecture 4 2D Fourier Transforms

Biophysical Techniques (BPHS 4090/PHYS 5800)

3. Image formation, Fourier analysis and CTF theory. Paula da Fonseca

Image restoration. Lecture 14. Milan Gavrilovic Centre for Image Analysis Uppsala University

Computational Aspects of MRI

Reading. 2. Fourier analysis and sampling theory. Required: Watt, Section 14.1 Recommended:

Reconstruction in CT and relation to other imaging modalities

LECTURE 3-1 AREA OF A REGION BOUNDED BY CURVES

ELEC Dr Reji Mathew Electrical Engineering UNSW

Optics Quiz #2 April 30, 2014

Image Restoration. Diffusion Denoising Deconvolution Super-resolution Tomographic Reconstruction

Index. aliasing artifacts and noise in CT images, 200 measurement of projection data, nondiffracting

Medical Image Reconstruction Term II 2012 Topic 6: Tomography

X-ray tomography. X-ray tomography. Applications in Science. X-Rays. Notes. Notes. Notes. Notes

Tutorial 5. Jun Xu, Teaching Asistant March 2, COMP4134 Biometrics Authentication

Topics. View Aliasing. CT Sampling Requirements. Sampling Requirements in CT Sampling Theory Aliasing

Contents. MATH 32B-2 (18W) (L) G. Liu / (TA) A. Zhou Calculus of Several Variables. 1 Homework 1 - Solutions 3. 2 Homework 2 - Solutions 13

CoE4TN4 Image Processing. Chapter 5 Image Restoration and Reconstruction

An Introduc+on to Mathema+cal Image Processing IAS, Park City Mathema2cs Ins2tute, Utah Undergraduate Summer School 2010

A Novel Two-step Method for CT Reconstruction

Refractive Index Map Reconstruction in Optical Deflectometry Using Total-Variation Regularization

Investigation on reconstruction methods applied to 3D terahertz computed Tomography

ange Data Reconstruction Using Fourier Slice Theorern*

Computer Vision I. Announcements. Fourier Tansform. Efficient Implementation. Edge and Corner Detection. CSE252A Lecture 13.

Fourier Transform in Image Processing. CS/BIOEN 6640 U of Utah Guido Gerig (slides modified from Marcel Prastawa 2012)

Lecture Image Enhancement and Spatial Filtering

Aliasing. Can t draw smooth lines on discrete raster device get staircased lines ( jaggies ):

EEE 512 ADVANCED DIGITAL SIGNAL AND IMAGE PROCESSING

Tomography. Forward projectionsp θ (r) are known as a Radon transform. Objective: reverse this process to form the original image

This is a good time to refresh your memory on double-integration. We will be using this skill in the upcoming lectures.

Steen Moeller Center for Magnetic Resonance research University of Minnesota

Corso di laurea in Fisica A.A Fisica Medica 4 TC

Exam 2 is Tue Nov 21. Bring a pencil and a calculator. Discuss similarity to exam1. HW3 is due Tue Dec 5.

Theoretically Perfect Sensor

Reconstruction in CT and relation to other imaging modalities

Sampling, Ordering, Interleaving

Theoretically Perfect Sensor

Visual Pathways to the Brain

Image Restoration. Yao Wang Polytechnic Institute of NYU, Brooklyn, NY 11201

Laser sensors. Transmitter. Receiver. Basilio Bona ROBOTICA 03CFIOR

Image Enhancement in Spatial Domain. By Dr. Rajeev Srivastava

Examination in Digital Image Processing, TSBB08

Material for Chapter 6: Basic Principles of Tomography M I A Integral Equations in Visual Computing Material

Filters and Fourier Transforms

Lecture note on Senior Laboratory in Phyhsics. Fraunhofer diffraction and double-split experiment

Temperature Distribution Measurement Based on ML-EM Method Using Enclosed Acoustic CT System

First CT Scanner. How it Works. Contemporary CT. Before and After CT. Computer Tomography: How It Works. Medical Imaging and Pattern Recognition

Camera Model and Calibration

Fourier transform of images

Non-Stationary CT Image Noise Spectrum Analysis

Drawing a Triangle (and an introduction to sampling)

Phys. 428, Lecture 6

MRI Physics II: Gradients, Imaging

Blind Sparse Motion MRI with Linear Subpixel Interpolation

Digital Image Fundamentals. Prof. George Wolberg Dept. of Computer Science City College of New York

Image Acquisition + Histograms

2D Fan Beam Reconstruction 3D Cone Beam Reconstruction

To Do. Advanced Computer Graphics. Sampling and Reconstruction. Outline. Sign up for Piazza

Lecture 6 Basic Signal Processing

Lecture 2 Image Processing and Filtering

Solutions For Homework #7

Important!!! First homework is due on Monday, September 26 at 8:00 am.

Image Reconstruction 3 Fully 3D Reconstruction

Transcription:

Multi-Dimensional Signals EE23 Digital Signal Processing Lecture Our world is more complex than D Images: f(x,y) Videos: f(x,y,t) Dynamic 3F scenes: f(x,y,z,t) Medical Imaging 3D Video Computer Graphics We will focus on 2D Continuous-Time 2D functions δ(x,y): Impulse at x=0, y=0 δ(x) : Impulse line (vertical or horizontal?) (x,y) : 2D rect function cos(2π(fxx + fyy) - Spatial harmonic Circularly Symmetic: δ(r): Impulse ring (x,y): Pillbox u(r) (r r 0 ) 2 r o Spatial Frequency What is a spatial frequency? Complex Harmonic: e j( xx+ y y) = e j2 (f xx+f y y) Units (for example): x,y - cm fx,fy - /cm Ωx, Ωy - rad/cm

Spatial Frequency Vinyl Record Transforms a temporal signal to a spatial signal Spatial Frequency CD ROM encodes digital temporal signals to spatial signals http://offtosognefjord.tumblr.com What is the frequency? What is the frequency? (-,) sin(2 (f x x + f y y)) (-,) sin(2 (f x x + f y y)) (-,-) a) fx=2, fy=2 b) fx=, fy=0 (,-) c) fx = 4, fy=0 d) none of the above (-,-) (,-) 2 cycles for 2 cm fx= cm -

What is the frequency? What is the frequency? sin(2 (f x x + f y y)) (-,) or cos(2 (f x x + f y y)) (-,) cos(2 (f x x + f y y)) (-,-) (,-) (-,-) (,-) a) sin, fx=0, fy=2 c) cos, fx = 0, fy=2 a) fx=, fy=2 c) fx = 2, fy= b) cos, fx=0, fy=4 d) none of the above b) fx=4, fy=2 d) fx=2, fy=4 What is the frequency? What is the Temporal Frequency? cos(2 (f x x + f y y)) Vinyl rotates at Hz (-,) (-,) (-,-) (,-) (-,-) (,-) a) fx=, fy=2 c) fx = 2, fy= a) cos(2π8t) c) cos(2π4t) b) fx=4, fy=2 d) fx=2, fy=4 b) cos(2π8t 2 ) d) cos(2π4t 2 )

What is the Temporal Frequency? (-,) Vinyl rotates at Hz Challenge: What is the password? Aliasing! 2sec (-,-) a) cos(2π00t) b) cos(2π00t 2 ) (,-) c) cos(2π40t) d) none of the answers 2D Fourier Transform Z Z F (f x,f y )= f(x, y)e j2 (f xx+f y y) dxdy

sinc(ax)sinc(by) 2D DTFT F (! x,! y )= X X f[n x,n y ]e j(! xn x +! y n y ) n x = n y = apple! x,! y apple sinc(ax)sinc(by) F (apple x, apple y )= I prefer 2nd X X n x = n y = f[n x,n y ]e j2 (apple xn x +apple y n y ) 0.5 apple apple x, apple y apple 0.5 Massaging the DTFT leads to separable transforms in each axis

Separability of 2D DTFT white board... 2D - DFT Similarly to D: Forward: F [k x,k y ]= Inverse: NX MX n x =0 n y =0 f[n x,n y ]e j2 (n xk x /N +n y k y /M ) apple x = k x /N, apple y = k y /M f[n x,n y ]= NM NX MX k x =0 k y =0 F [k x,k y ]e +j2 (n xk x /N +n y k y /M ) 2D - DFT F [k x,k y ]= M NX MX n x =0 n y =0 f[n x,n y ]e j2 (n xk x /N +n y k y /M ) f[n x,n y ] F [k x,k y ] (0,) Properties of 2D DFT Circular Convolution f[n x,n y ] h[n x,n y ]=F [k x,k y ]H[k x,k y ] ny ky Circular shift (0,0) N nx (0,0) kx (,0) Need to fftshift in 2D to get it to look like DTFT. f[(n x m x ) N, (n y m y ) M ]=e j2 (k xm x /N +k y m y /M ) F [k x,k y ]

. Projection w Many Projections - Tomography (A fr:x,j) J(f-xco,e-j>l'ne)rfx:dj / \:. J projection x-ray http://www.youtube.com/watch?v=4gklqhm9ay&feature=related

/ \:. J. Radon Transform w Radon Transform: Sinogram (A fr:x,j) J(f-xco,e-j>l'ne)rfx:dj / \:. J Computed Tomography Projection Slice Theorem (Bracewell) Sinogram cross-section,f F0<. (), ey'" fy{) FBP M. lrd ft spatial frequency x-ray source central-section/ projection slice theorem (Bracewell 56)?voo P: o-tu : Ar fcx,j)dj F( /.;_,o) =- JJ G "' o ='l k, or6f 9 /J-fr{ t,:x-.. f; jj d-xjj =-

Projection Slice Theorem (Bracewell) Discrete Reconstruction?voo P : o-tu : A r G "' o ='l k, or6f 9.. fcx,j)dj F( /.;_,o) =- J J /J-fr{ t,:x-.. f; jj d-xjj =- -\A - (.)0..... ' Discretization /)L{c(2EJ7.7:.[.'. - -- -