# Digital image processing

Save this PDF as:

Size: px
Start display at page:

## Transcription

1 Digital image processing The two-dimensional discrete Fourier transform and applications: image filtering in the frequency domain Introduction Frequency domain filtering modifies the brightness values of the pixels in the image depending on their periodic patterns and on the spatial distributions of brightness variations in the image. The high pass filters tend to preserve the sharp brightness variation patterns and to remove the smooth variations in the image (along with the DC component, i.e. with the average brightness which is set to zero). The low pass filters, on the opposite, preserve the smooth brightness variations in the image, and remove the sharp edges (sharp brightness variations) in the image. To apply a frequency domain filtering procedure, the most typical approach is to first apply a Fourier transform to the image, to obtain its frequency domain representation; the frequency domain representation of the image through its Fourier transform provides information about the spectral components of the image. The frequency domain filtering process simply modifies certain spectral components in the image and afterwards, by applying the inverse Fourier transform on the frequency domain filtered image represented in the frequency domain, provides the filtered image representation in the spatial domain. Filtering Inverse FFT FFT f(x,y) F(u,v) G(u,v) g(x,y) Figure The architecture of a frequency domain filtering system In a digital image, the details and the sharp boundaries are associated to high spatial frequencies, since they introduce sharp brightness transitions. On the opposite, the approximately uniform surfaces and background areas are associated to the low spatial frequencies, since they correspond to the smooth variations of the brightness or even to no spatial variations of the brightness. A digital image can exhibit, e.g., a periodic noise in the form of some periodical lines, resulting from an imperfect digitization. In the spatial frequency domain, this noise will introduce spectral components in the high frequency range, with a good localization due to its periodicity. If one filters out these spectral components by setting the corresponding coefficients of the Fourier transform to zero, and afterwards applies the inverse Fourier transform on the modified spectrum, the result is an image without noise and typically without any noticeable loss in details. The Discrete Fourier Transform The discrete Fourier transform is one of the most important transforms used in signal processing and image processing. For a -D periodical sequence of samples {u(n), n=,...,-}, its discrete Fourier transform is given by:

2 7. The -D DFT and applications: image filtering in the frequency domain - kn n= v(k)= u(n), k =,.., - with: -jπ = exp( ) The inverse discrete Fourier transform of this sequence is: u(n) = - kn v(k) ; n =,,..., - k= The two-dimensional discrete Fourier transform The -D discrete Fourier transform of an image of size M pixels, U[Mx]={u(m,n)}, is a separable transform defined by the expression: v(k,l)= u(m,n), k M - l - M M- - km ln M ; 3 m= n= The inverse -D discrete Fourier transform is defined as: M - - u(m,n)= v( k, l) km ln M, m M - ; n - 4 k= l= Let s consider the computation of the -D DFT of an image of size 3 3 pixels, whose isometric (3-D) representation is given in Figure (left) the image is a square of size pixels, centred in the image; the background of the image is black, and the square is white (and for convenience, the brightness is normalized in the range [;]). The -D DFT amplitude spectrum of this image is shown in Figure in the centre; the right diagram in Figure represents the amplitude spectrum shifted by mirroring to give the DC component in the centre and the high frequencies on the corners. The digital image represented in 3D The amplitude spectrum Shifted amplitude spectrum Figure Exemplification of the -D DFT Interpretation of the DFT Using trigonometry relations, one can re-write the inverse -D DFT in the following form: / π u f ( x) = F () + F ( u ) cos( + phase ( F ( u )) 5 u = This expression shows that the original signal can be expressed as a weighted sum of cosine basis functions. Approximating this sum by only the first two largest terms, we get the following approximation of the signal:

3 Digital image processing π + o π 3 f ( x) cos, phase : 9 +. cos, phase : 9 o 6 The horizontal profile of the reconstructed signal is the following: Fast algorithms for the discrete Fourier transform I. Rows and columns decomposition Since the DFT is extensively used in signal processing, there was a great interest in developing algorithms for its efficient computation (also taking into account that DFT may be used in computing other signal/image transforms, such as DCT). To estimate the efficiency of the computation of -D DFT of a digital image by the rows and columns decomposition method, let us first examine the computational complexity of the direct computation of the -D DFT for a complex -D sequence of points, denoted by x(n,n ). The -D DFT for this sequence is denoted by X ( k, k ). The graphical representation of the rows and columns decomposition is given in Figure 3: n ( -) k (-) D-DFT in points ( -) n ( -) k 3

4 7. The -D DFT and applications: image filtering in the frequency domain n ( -) k ( -) D-DFT in points ( -) n ( -) k Figure 3 Illustration of the rows and columns decomposition in computing DFT The result of the application of this algorithm on some digital image is shown in Figure 4: Original input image DFT applied on the rows DFT applied on the columns Figure 4 Example of applying the DFT D by the rows and columns decomposition method According to the above considerations, the -D DFT sequence X ( k, k ) may be computed by a total of -D DFTs applied on a series of D-DFTs. Assuming the computation of the -D DFTs according to the Equation above (which implies multiplications and additions), the total number of arithmetic operations involved in the computation of the sequence X is (+) multiplications and additions. Therefore the number of operations is much smaller as compared to the direct computation given by the Equation 3. II. The RADIX- FFT algorithm One of the most efficient algorithms in the computation of the DFT is based on the divide and conquer technique. This algorithm is only applicable when the length of the sequence,, is not prime, i.e. when can be decomposed as a product of integers =r*r*...*rx, where {rj} are primes. A particular case appears for r=r=...=rx=r; in this case = r x, and the DFT of the sequence may be decomposed in a set of DFTs applied to some sequences of length r. The value r is called the radix of the FFT algorithm. Let us consider the computation of the DFT of a sequence of length = v samples through the divide and conquer method. If one denotes by M=/ and L=, the initial sequence of samples may be divided into two vectors, corresponding to the even and odd indices in the sequence. The two sub-sequences are obtained by the decimation of the initial sequence x(n) by a factor of ; therefore the FFT algorithm in this case is called time decimation FFT algorithm. 4

5 Digital image processing f ( n) = x( n) f ( n) = x( n + ) n =,... 7 ow one can express the DFT of the sequence of length in terms of the DFTs of the decimated sequences as follows: X ( k ) = = = n= x( n) n= par / m= kn x( n) kn x(m) + mk k =,,..., n= impar + x( n) / m= kn x(m + ) (m+ ) k But - jπ = exp ( ), therefore =, which leads to the following / reformulation of the above expression: / / mk k X ( k) = f( m) / + f( m) m= m= k = F ( k) + F ( k) where F ( k ) and F k ( ) are the DFTs of the two (even and odd) sub-sequences of length /: f and f. If F (k) and F (k) are periodical with the period /, then F ( k + / ) = F ( k ) and F ( k + / ) = F ( k ) ; furthermore, considering that + / =, one can write: X( k) = F ( k) + X( k + ) = F ( k) k F ( k) k F ( k) mk / k k k =,,..., k =,,..., The direct computation of F ( k ) requires /4 complex multiplications (and the same number is needed for F ( k ) ). Other / complex multiplications are needed for the term k F ( k). Therefore the computation of X(k) requires + = + complex multiplications. Thus, even after this first step, one can estimate a computational complexity reduction, from to only +, which is approximately equivalent to a reduction by a factor of (approximately) for large. Denoting by: G(k) = F ( k) k =,,..., k G (k) = F ( k) k =,,..., 9 5

6 7. The -D DFT and applications: image filtering in the frequency domain the above algorithm can be repeated in a recursive fashion, leading to the radix FFT. To illustrate the operation of this algorithm, let us consider the particular case of a sequence of length =. In general, each butterfly implies one complex multiplication and two complex additions. For an = ν, / butterflies and v = log stages are needed; these will require a number of operations derived above. A butterfly implies one multiply and two additions between the pair of complex numbers (a,b) to give as output the pair (A,B), and there is no need to store the input pair (a,b). Therefore the output result of the butterfly computation (A,B) may be stored in the same location as the input (a,b), which is very efficient from the memory requirements point of view. The basic butterfly corresponding to the time decimation is: a A=a+ r *b - b B=a - r *b Basic butterfly To understand the workflow of the algorithm, one should examine the way the data is updated in the memory, after each decimation stage. This is illustrated in Figure 5: x() x() x() x(3) x(4) x(5) x(6) x(7) x() x() x(4) x(6) x() x(3) x(5) x(7) decimation Data reordering decimation x() x(4) x() x(6) x() x(5) x(3) x(7) Figure 5 The decimation algorithm The previous figure suggests that, whereas the input data are in their normal order, the output data are in a reversed order. This issue must be taken into account in the frequency domain filtering algorithms, in order to appropriately filter the desired spectral components. The entire FFT procedure is presented in Figure 6. This graph is useful to understand the disadvantages of the algorithm for large values of the most significant being the following: - the transform coefficients for large values of take large values as well, therefore their handling is difficult (in respect to the memory requirements); - considering the large number of multiplications needed, it becomes clear that the errors accumulation cannot be neglected for large values of ; these errors will reflect on the reconstructed image, even at no truncation of the transform coefficients. 6

7 Digital image processing x() X() x() - X(4) x() - X() x(3) - - X(6) x(4) - X() x(5) - - X(5) x(6) - - X(3) x(7) X(7) Figure 6 The FFT graph in the time decimation algorithm, for = Displaying the Discrete Fourier Transform The following four components describing complex numbers may be used in the visualization of an FFT image: the real part; the imaginary part; the amplitude (absolute value); the phase. The definition of these four components is described by the following expressions: (, ) (, ) (, ) F u v = R u v + ji u v where R(u,v) is the real part, and I(u,v) is the imaginary part, or: j ( u, v) (, ) (, ) F u v = F u v e ϕ 3 where F(u,v) is the amplitude, and ϕ(u,v) is the phase. The amplitude F(u,v) is also called the Fourier spectrum, defined as: (, ) (, ) (, ) = + 4 F u v R u v I u v The square of the Fourier spectrum is called Fourier power spectrum or spectral density. The phase ϕ(u,v) is also called phase angle and is defined as: 7

8 7. The -D DFT and applications: image filtering in the frequency domain (, ) (, ) I u v ϕ ( u, v) = arctan R u v Consider an image with M pixels, and denote by x and y the horizontal and vertical sampling steps. Then the Fourier transform of this image will have the same size, M, and the sampling frequencies u and v will be defined by the expressions: u = ; v = x M y There are two ways to represent the Fourier spectrum of a digital image: the default representation the optical representation The default representation In the case of the standard (default) representation (Figure 7), the high frequencies are grouped in the centre of the Fourier spectrum image, whereas the low frequencies are located on the corners of this image. Thus, the zero frequency (the DC coefficient) is, u, M v located in the four corners of the image. The frequency range is: [ ] [ ] 5 6 A Low frequency B High frequency C D The optical representation Figure 7 The standard (default) representation In the case of the optical representation (Figure ), the low frequencies are located in the centre of the Fourier spectrum image, and the high frequencies are located in the corners of the Fourier spectrum image. Thus, the zero frequency (the DC coefficient) is located in the centre of the Fourier spectrum image. The frequency range is: M M u, u v, v

9 Digital image processing A High frequency B Low frequency C D Figure The optical representation Frequency domain filtering The low pass filtering A frequency domain low pass filter attenuates or sets to zero the high frequency components in the Fourier spectrum. Such a filter suppresses the information introduced by the sharp spatial variations of the brightness in the digital image. As a result of applying such a filter on a digital image, the reconstructed image (through the inverse Fourier transform) has less noise, but the details and the contours of the objects as well as the textures of the surfaces in the image are also attenuated. Low pass attenuation filters In this case, the spectral components are linearly reduced as amplitude (see Figure 9.a)), starting from the DC component (of frequency f ) to the maximum frequency component (of frequency f max ). The algorithm implies the multiplication of all the spectral components by a coefficient C(f) given by the expression: C f f max ( f ) = with C(f )= and C(f max )=. f f max 7 C(f) C(f) f f max f f c f max a) b) Figure 9 The low pass attenuation (a) and low pass truncation (b) 9

10 7. The -D DFT and applications: image filtering in the frequency domain Low pass truncation filters The low pass truncation filters (Figure 9.b)) simply eliminates all the frequency components whose frequencies f are higher than some cut off frequency f c. This is done by multiplying each frequency component f by the coefficient C, taking the values {,}: The high pass filtering ; if f > f C ( f ) = c ; otherwise A frequency domain high pass filter attenuates or sets to zero the low frequency components in the Fourier spectrum. Such a filter suppresses the information introduced by the smooth spatial variations of the brightness in the digital image. As a result of applying such a filter on a digital image, the reconstructed image (through the inverse Fourier transform) has its uniform or quasi-uniform brightness set to zero. High pass attenuation filters In this case, the spectral components are linearly reduced as amplitude with a positive slope (see Figure.a)), starting from the DC component (of frequency f ) to the maximum frequency component (of frequency f max ). The algorithm implies the multiplication of all the spectral components by a coefficient C(f) given by the expression: with C(f )= and C(f max )=. C f f f f ( f ) = max 9 C(f) C(f) f f max f f c f max a) b) Figure The high pass attenuation (a) and the high pass truncation (b) High pass truncation filters The high pass truncation filters (Figure 9.b)) simply eliminates all the frequency components whose frequencies f are lower than some cut off frequency f c. This is done by multiplying each frequency component f by the coefficient C, taking the values {,}: ; dacã f < f C ( f ) = c ; altfel

11 LabView IMAQ Functions for DFT Applications IMAQ GetPalette Selects a display palette. Five predefined palettes are available. To activate a color palette choose a code for Palette umber and connect the Color Palette output to the input Color Palette of IMAQ inddraw. Digital image processing Palette umber (gray) gives you a choice of five predefined palettes. You can choose from the following values: Gray - Grayscale is the default palette. Binary - designed especially for binary images. Gradient Rainbow Temperature Color Palette indicates an array of clusters composed of 56 elements for each of the three color planes. IMAQ FFT Computes the FFT of an image. The FFT is a complex image in which high frequencies are grouped at the center, while low frequencies are located at the corners. Image Src - is the handle of the source image. Image Dst - is the handle of the complex image that contains the resulting FFT image. This input can accept only a complex image. Image Dst Out - is the reference to the destination (output) image that receives the processing results of the VI. IMAQ InverseFFT Computes the inverse FFT of a complex image ( 3-bit floating point). Image Src - is the handle of the source image. This input can accept only a complex image. Image Dst - is the handle of the -bit, 6-bit, or 3-bit floating-point image that contains the resulting spatial image. Image Dst Out - is the reference to the destination (output) image that receives the processing results of the VI. IMAQ ComplexTruncate Truncates the frequencies of a complex image.

12 7. The -D DFT and applications: image filtering in the frequency domain Low pass/high pass (Low pass) - determines which frequencies are truncated. Choose Low pass (F) to remove the high frequencies or High pass (T) to remove the low frequencies. The default is FALSE, which specifies lowpass. Truncation Frequency % - is the percentage of the frequencies that are retained within a Fourier-transformed image. Image Src - is the image reference source. It must be a complex image. Image Dst - is the reference of the image destination. If it is connected, it must be the same type as the Image Src. Image Dst Out - is the reference to the destination (output) image that receives the processing results of the VI. Practice Fourier transform implementation for a greyscale image open a new LabView session create a new project in the diagram window, add the IMAQ functions needed to obtain the following schematic:

13 Digital image processing the user interface should be the following: load different images, perform the FFT and notice the distribution of the FFT coefficients. Modify the color palette used to display the Fourier coefficients and observe the effect on the display of these images. Locate the low frequency, mean frequency and high frequency regions in the Fourier spectrum. Frequency domain filtering for a greyscale image open a new LabView session create a new project in the diagram window, add the IMAQ functions needed to obtain the following schematic: 3

14 7. The -D DFT and applications: image filtering in the frequency domain the user interface should be the following: load different images, perform LPF and HPF for different filtering percentages (see the description of the IMAQ Complex Truncate function). Modify the 4

15 Digital image processing colour palette used for display of the FFT coefficients and observe the effect on the display window draw a block diagram to perform a band pass filtering, where the both thresholds, inferior and superior one, may be modified. Questions and exercises. Implement a diagram that will perform the Fourier transform of a RGB color image.. Modify the first block diagram to obtain a 3D representation of the Fourier spectrum, before and after filtering. 3. Implement a LabView application which will perform highpass attenuation and lowpass attenuation for a grayscale image 5

### Digital Image Processing

1 Introduction in IMAQ Vision. Structure of Digital Image Processing Applications Using IMAQ Vision 1.1 Introduction A grey level image is a function expressing the spatial variation of the light intensity

### FFT Algorithms. Chapter 6. Contents 6.1

Chapter 6 FFT Algorithms Contents Efficient computation of the DFT............................................ 6.2 Applications of FFT................................................... 6.6 Computing DFT

### Image Enhancement: Frequency domain methods

Image Enhancement: Frequency domain methods The concept of filtering is easier to visualize in the frequency domain. Therefore, enhancement of image f ( m, n) can be done in the frequency domain, based

### Fast Fourier Transforms and Power Spectra in LabVIEW

Application Note 4 Introduction Fast Fourier Transforms and Power Spectra in LabVIEW K. Fahy, E. Pérez Ph.D. The Fourier transform is one of the most powerful signal analysis tools, applicable to a wide

### Frequency domain filtering fundamentals

Frequency domain filtering fundamentals by Gleb V. Tcheslavski: gleb@ee.lamar.edu http://ee.lamar.edu/gleb/dip/index.htm Spring 2008 ELEN 4304/5365 DIP 1 Preliminaries For a digital image f(x,y) the basic

### Polynomials and the Fast Fourier Transform (FFT) Battle Plan

Polynomials and the Fast Fourier Transform (FFT) Algorithm Design and Analysis (Wee 7) 1 Polynomials Battle Plan Algorithms to add, multiply and evaluate polynomials Coefficient and point-value representation

### Image Processing with. ImageJ. Biology. Imaging

Image Processing with ImageJ 1. Spatial filters Outlines background correction image denoising edges detection 2. Fourier domain filtering correction of periodic artefacts 3. Binary operations masks morphological

### The continuous and discrete Fourier transforms

FYSA21 Mathematical Tools in Science The continuous and discrete Fourier transforms Lennart Lindegren Lund Observatory (Department of Astronomy, Lund University) 1 The continuous Fourier transform 1.1

### Moving Average Filters

CHAPTER 15 Moving Average Filters The moving average is the most common filter in DSP, mainly because it is the easiest digital filter to understand and use. In spite of its simplicity, the moving average

### ALFFT FAST FOURIER Transform Core Application Notes

ALFFT FAST FOURIER Transform Core Application Notes 6-20-2012 Table of Contents General Information... 3 Features... 3 Key features... 3 Design features... 3 Interface... 6 Symbol... 6 Signal description...

### SIGNAL PROCESSING & SIMULATION NEWSLETTER

1 of 10 1/25/2008 3:38 AM SIGNAL PROCESSING & SIMULATION NEWSLETTER Note: This is not a particularly interesting topic for anyone other than those who ar e involved in simulation. So if you have difficulty

### jorge s. marques image processing

image processing images images: what are they? what is shown in this image? What is this? what is an image images describe the evolution of physical variables (intensity, color, reflectance, condutivity)

### Fourier Transform and Image Filtering. CS/BIOEN 6640 Lecture Marcel Prastawa Fall 2010

Fourier Transform and Image Filtering CS/BIOEN 6640 Lecture Marcel Prastawa Fall 2010 The Fourier Transform Fourier Transform Forward, mapping to frequency domain: Backward, inverse mapping to time domain:

### Lecture 14. Point Spread Function (PSF)

Lecture 14 Point Spread Function (PSF), Modulation Transfer Function (MTF), Signal-to-noise Ratio (SNR), Contrast-to-noise Ratio (CNR), and Receiver Operating Curves (ROC) Point Spread Function (PSF) Recollect

### Computer Vision & Digital Image Processing

Intensity Transformations and Spatial Filtering Basics Computer Vision & Digital Image Processing Intensity Transformations and Spatial Filtering Operations take place in the spatial domain Operate directly

### Digital Image Processing. Prof. P.K. Biswas. Department of Electronics & Electrical Communication Engineering

Digital Image Processing Prof. P.K. Biswas Department of Electronics & Electrical Communication Engineering Indian Institute of Technology, Kharagpur Lecture - 27 Colour Image Processing II Hello, welcome

### Lectures 6&7: Image Enhancement

Lectures 6&7: Image Enhancement Leena Ikonen Pattern Recognition (MVPR) Lappeenranta University of Technology (LUT) leena.ikonen@lut.fi http://www.it.lut.fi/ip/research/mvpr/ 1 Content Background Spatial

### FAST Fourier Transform (FFT) and Digital Filtering Using LabVIEW

FAST Fourier Transform (FFT) and Digital Filtering Using LabVIEW Wei Lin Department of Biomedical Engineering Stony Brook University Instructor s Portion Summary This experiment requires the student to

### Performing the Fast Fourier Transform with Microchip s dspic30f Series Digital Signal Controllers

Performing the Fast Fourier Transform with Microchip s dspic30f Series Digital Signal Controllers Application Note Michigan State University Dept. of Electrical & Computer Engineering Author: Nicholas

### Analog and Digital Signals, Time and Frequency Representation of Signals

1 Analog and Digital Signals, Time and Frequency Representation of Signals Required reading: Garcia 3.1, 3.2 CSE 3213, Fall 2010 Instructor: N. Vlajic 2 Data vs. Signal Analog vs. Digital Analog Signals

### 1.4 Fast Fourier Transform (FFT) Algorithm

74 CHAPTER AALYSIS OF DISCRETE-TIME LIEAR TIME-IVARIAT SYSTEMS 4 Fast Fourier Transform (FFT Algorithm Fast Fourier Transform, or FFT, is any algorithm for computing the -point DFT with a computational

### Removal of Noise from MRI using Spectral Subtraction

International Journal of Electronic and Electrical Engineering. ISSN 0974-2174, Volume 7, Number 3 (2014), pp. 293-298 International Research Publication House http://www.irphouse.com Removal of Noise

### VLSI Architecture for DCT Based On High Quality DA

International Journal of Engineering and Technical Research (IJETR) ISSN: 2321-0869, Volume-2, Issue-6, June 2014 VLSI Architecture for DCT Based On High Quality DA Urbi Sharma, Tarun Verma, Rita Jain

### APPLICATION OF FILTER BANK THEORY TO SUBBAND CODING OF IMAGES

EC 623 ADVANCED DIGITAL SIGNAL PROCESSING TERM-PROJECT APPLICATION OF FILTER BANK THEORY TO SUBBAND CODING OF IMAGES Y. PRAVEEN KUMAR 03010240 KANCHAN MISHRA 03010242 Supervisor: Dr. S.R.M. Prasanna Department

### DIGITAL IMAGE PROCESSING AND ANALYSIS

DIGITAL IMAGE PROCESSING AND ANALYSIS Human and Computer Vision Applications with CVIPtools SECOND EDITION SCOTT E UMBAUGH Uffi\ CRC Press Taylor &. Francis Group Boca Raton London New York CRC Press is

### Admin stuff. 4 Image Pyramids. Spatial Domain. Projects. Fourier domain 2/26/2008. Fourier as a change of basis

Admin stuff 4 Image Pyramids Change of office hours on Wed 4 th April Mon 3 st March 9.3.3pm (right after class) Change of time/date t of last class Currently Mon 5 th May What about Thursday 8 th May?

### TI-86 Graphing Calculator Keystroke Guide

TI-86 Graphing Calculator Keystroke Guide In your textbook you will notice that on some pages a key-shaped icon appears next to a brief description of a feature on your graphing calculator. In this guide

### Comparison of different image compression formats. ECE 533 Project Report Paula Aguilera

Comparison of different image compression formats ECE 533 Project Report Paula Aguilera Introduction: Images are very important documents nowadays; to work with them in some applications they need to be

### Sachin Patel HOD I.T Department PCST, Indore, India. Parth Bhatt I.T Department, PCST, Indore, India. Ankit Shah CSE Department, KITE, Jaipur, India

Image Enhancement Using Various Interpolation Methods Parth Bhatt I.T Department, PCST, Indore, India Ankit Shah CSE Department, KITE, Jaipur, India Sachin Patel HOD I.T Department PCST, Indore, India

### 3. Interpolation. Closing the Gaps of Discretization... Beyond Polynomials

3. Interpolation Closing the Gaps of Discretization... Beyond Polynomials Closing the Gaps of Discretization... Beyond Polynomials, December 19, 2012 1 3.3. Polynomial Splines Idea of Polynomial Splines

### Digital Image Processing

Digital Image Processing Using MATLAB Second Edition Rafael C. Gonzalez University of Tennessee Richard E. Woods MedData Interactive Steven L. Eddins The MathWorks, Inc. Gatesmark Publishing A Division

### The Algorithms of Speech Recognition, Programming and Simulating in MATLAB

FACULTY OF ENGINEERING AND SUSTAINABLE DEVELOPMENT. The Algorithms of Speech Recognition, Programming and Simulating in MATLAB Tingxiao Yang January 2012 Bachelor s Thesis in Electronics Bachelor s Program

### Grade 6 Mathematics Performance Level Descriptors

Limited Grade 6 Mathematics Performance Level Descriptors A student performing at the Limited Level demonstrates a minimal command of Ohio s Learning Standards for Grade 6 Mathematics. A student at this

### RN-Codings: New Insights and Some Applications

RN-Codings: New Insights and Some Applications Abstract During any composite computation there is a constant need for rounding intermediate results before they can participate in further processing. Recently

### Mathematical Procedures

CHAPTER 6 Mathematical Procedures 168 CHAPTER 6 Mathematical Procedures The multidisciplinary approach to medicine has incorporated a wide variety of mathematical procedures from the fields of physics,

### The Fourier Analysis Tool in Microsoft Excel

The Fourier Analysis Tool in Microsoft Excel Douglas A. Kerr Issue March 4, 2009 ABSTRACT AD ITRODUCTIO The spreadsheet application Microsoft Excel includes a tool that will calculate the discrete Fourier

### Lab 1. The Fourier Transform

Lab 1. The Fourier Transform Introduction In the Communication Labs you will be given the opportunity to apply the theory learned in Communication Systems. Since this is your first time to work in the

### Wavelet analysis. Wavelet requirements. Example signals. Stationary signal 2 Hz + 10 Hz + 20Hz. Zero mean, oscillatory (wave) Fast decay (let)

Wavelet analysis In the case of Fourier series, the orthonormal basis is generated by integral dilation of a single function e jx Every 2π-periodic square-integrable function is generated by a superposition

### THE FAST FOURIER TRANSFORM

THE FAST FOURIER TRASFORM LOG CHE ABSTRACT. Fast Fourier transform (FFT) is a fast algorithm to compute the discrete Fourier transform in O( log ) operations for an array of size 2 J. It is based on the

### L9: Cepstral analysis

L9: Cepstral analysis The cepstrum Homomorphic filtering The cepstrum and voicing/pitch detection Linear prediction cepstral coefficients Mel frequency cepstral coefficients This lecture is based on [Taylor,

### Aliasing, Image Sampling and Reconstruction

Aliasing, Image Sampling and Reconstruction Recall: a pixel is a point It is NOT a box, disc or teeny wee light It has no dimension It occupies no area It can have a coordinate More than a point, it is

### RN-coding of Numbers: New Insights and Some Applications

RN-coding of Numbers: New Insights and Some Applications Peter Kornerup Dept. of Mathematics and Computer Science SDU, Odense, Denmark & Jean-Michel Muller LIP/Arénaire (CRNS-ENS Lyon-INRIA-UCBL) Lyon,

### Solve addition and subtraction word problems, and add and subtract within 10, e.g., by using objects or drawings to represent the problem.

Solve addition and subtraction word problems, and add and subtract within 10, e.g., by using objects or drawings to represent the problem. Solve word problems that call for addition of three whole numbers

### Color image processing: pseudocolor processing

Color image processing: pseudocolor processing by Gleb V. Tcheslavski: gleb@ee.lamar.edu http://ee.lamar.edu/gleb/dip/index.htm Spring 2008 ELEN 4304/5365 DIP 1 Preliminaries Pseudocolor (false color)

### Matlab GUI for WFB spectral analysis

Matlab GUI for WFB spectral analysis Jan Nováček Department of Radio Engineering K13137, CTU FEE Prague Abstract In the case of the sound signals analysis we usually use logarithmic scale on the frequency

### Final Year Project Progress Report. Frequency-Domain Adaptive Filtering. Myles Friel. Supervisor: Dr.Edward Jones

Final Year Project Progress Report Frequency-Domain Adaptive Filtering Myles Friel 01510401 Supervisor: Dr.Edward Jones Abstract The Final Year Project is an important part of the final year of the Electronic

### Fast Fourier Transform: Theory and Algorithms

Fast Fourier Transform: Theory and Algorithms Lecture Vladimir Stojanović 6.973 Communication System Design Spring 006 Massachusetts Institute of Technology Discrete Fourier Transform A review Definition

### 4.3 Analog-to-Digital Conversion

4.3 Analog-to-Digital Conversion overview including timing considerations block diagram of a device using a DAC and comparator example of a digitized spectrum number of data points required to describe

### Frequency Response and Continuous-time Fourier Transform

Frequency Response and Continuous-time Fourier Transform Goals Signals and Systems in the FD-part II I. (Finite-energy) signals in the Frequency Domain - The Fourier Transform of a signal - Classification

### Utah Core Curriculum for Mathematics

Core Curriculum for Mathematics correlated to correlated to 2005 Chapter 1 (pp. 2 57) Variables, Expressions, and Integers Lesson 1.1 (pp. 5 9) Expressions and Variables 2.2.1 Evaluate algebraic expressions

### Frequency Domain Characterization of Signals. Yao Wang Polytechnic University, Brooklyn, NY11201 http: //eeweb.poly.edu/~yao

Frequency Domain Characterization of Signals Yao Wang Polytechnic University, Brooklyn, NY1121 http: //eeweb.poly.edu/~yao Signal Representation What is a signal Time-domain description Waveform representation

### Department of Electrical and Computer Engineering Ben-Gurion University of the Negev. LAB 1 - Introduction to USRP

Department of Electrical and Computer Engineering Ben-Gurion University of the Negev LAB 1 - Introduction to USRP - 1-1 Introduction In this lab you will use software reconfigurable RF hardware from National

### Analysis/resynthesis with the short time Fourier transform

Analysis/resynthesis with the short time Fourier transform summer 2006 lecture on analysis, modeling and transformation of audio signals Axel Röbel Institute of communication science TU-Berlin IRCAM Analysis/Synthesis

### Direct and Reflected: Understanding the Truth with Y-S 3

Direct and Reflected: Understanding the Truth with Y-S 3 -Speaker System Design Guide- December 2008 2008 Yamaha Corporation 1 Introduction Y-S 3 is a speaker system design software application. It is

### Common Core Unit Summary Grades 6 to 8

Common Core Unit Summary Grades 6 to 8 Grade 8: Unit 1: Congruence and Similarity- 8G1-8G5 rotations reflections and translations,( RRT=congruence) understand congruence of 2 d figures after RRT Dilations

### Natural Neighbour Interpolation

Natural Neighbour Interpolation DThe Natural Neighbour method is a geometric estimation technique that uses natural neighbourhood regions generated around each point in the data set. The method is particularly

### Segmentation and Automatic Descreening of Scanned Documents

Segmentation and Automatic Descreening of Scanned Documents Alejandro Jaimes a, Frederick Mintzer b, A. Ravishankar Rao b and Gerhard Thompson b a Columbia University b IBM T.J. Watson Research Center

### chapter Introduction to Digital Signal Processing and Digital Filtering 1.1 Introduction 1.2 Historical Perspective

Introduction to Digital Signal Processing and Digital Filtering chapter 1 Introduction to Digital Signal Processing and Digital Filtering 1.1 Introduction Digital signal processing (DSP) refers to anything

### TYPES OF NUMBERS. Example 2. Example 1. Problems. Answers

TYPES OF NUMBERS When two or more integers are multiplied together, each number is a factor of the product. Nonnegative integers that have exactly two factors, namely, one and itself, are called prime

### Personal Identity Verification (PIV) IMAGE QUALITY SPECIFICATIONS FOR SINGLE FINGER CAPTURE DEVICES

Personal Identity Verification (PIV) IMAGE QUALITY SPECIFICATIONS FOR SINGLE FINGER CAPTURE DEVICES 1.0 SCOPE AND PURPOSE These specifications apply to fingerprint capture devices which scan and capture

### SOFTWARE FOR GENERATION OF SPECTRUM COMPATIBLE TIME HISTORY

3 th World Conference on Earthquake Engineering Vancouver, B.C., Canada August -6, 24 Paper No. 296 SOFTWARE FOR GENERATION OF SPECTRUM COMPATIBLE TIME HISTORY ASHOK KUMAR SUMMARY One of the important

### CHAPTER 3: DIGITAL IMAGING IN DIAGNOSTIC RADIOLOGY. 3.1 Basic Concepts of Digital Imaging

Physics of Medical X-Ray Imaging (1) Chapter 3 CHAPTER 3: DIGITAL IMAGING IN DIAGNOSTIC RADIOLOGY 3.1 Basic Concepts of Digital Imaging Unlike conventional radiography that generates images on film through

### Sampling Theorem Notes. Recall: That a time sampled signal is like taking a snap shot or picture of signal periodically.

Sampling Theorem We will show that a band limited signal can be reconstructed exactly from its discrete time samples. Recall: That a time sampled signal is like taking a snap shot or picture of signal

### PIXEL-LEVEL IMAGE FUSION USING BROVEY TRANSFORME AND WAVELET TRANSFORM

PIXEL-LEVEL IMAGE FUSION USING BROVEY TRANSFORME AND WAVELET TRANSFORM Rohan Ashok Mandhare 1, Pragati Upadhyay 2,Sudha Gupta 3 ME Student, K.J.SOMIYA College of Engineering, Vidyavihar, Mumbai, Maharashtra,

### Big Ideas in Mathematics

Big Ideas in Mathematics which are important to all mathematics learning. (Adapted from the NCTM Curriculum Focal Points, 2006) The Mathematics Big Ideas are organized using the PA Mathematics Standards

### Mathematics. Mathematical Practices

Mathematical Practices 1. Make sense of problems and persevere in solving them. 2. Reason abstractly and quantitatively. 3. Construct viable arguments and critique the reasoning of others. 4. Model with

### HSI BASED COLOUR IMAGE EQUALIZATION USING ITERATIVE n th ROOT AND n th POWER

HSI BASED COLOUR IMAGE EQUALIZATION USING ITERATIVE n th ROOT AND n th POWER Gholamreza Anbarjafari icv Group, IMS Lab, Institute of Technology, University of Tartu, Tartu 50411, Estonia sjafari@ut.ee

### ANALYZER BASICS WHAT IS AN FFT SPECTRUM ANALYZER? 2-1

WHAT IS AN FFT SPECTRUM ANALYZER? ANALYZER BASICS The SR760 FFT Spectrum Analyzer takes a time varying input signal, like you would see on an oscilloscope trace, and computes its frequency spectrum. Fourier's

### Note 5.1 Stress range histories and Rain Flow counting

June 2009/ John Wægter Note 5.1 Stress range histories and Rain Flow counting Introduction...2 Stress range histories...3 General...3 Characterization of irregular fatigue loading...4 Stress range histogram...5

### Digital image processing

746A27 Remote Sensing and GIS Lecture 4 Digital image processing Chandan Roy Guest Lecturer Department of Computer and Information Science Linköping University Digital Image Processing Most of the common

### An Energy-Based Vehicle Tracking System using Principal Component Analysis and Unsupervised ART Network

Proceedings of the 8th WSEAS Int. Conf. on ARTIFICIAL INTELLIGENCE, KNOWLEDGE ENGINEERING & DATA BASES (AIKED '9) ISSN: 179-519 435 ISBN: 978-96-474-51-2 An Energy-Based Vehicle Tracking System using Principal

### Video-Conferencing System

Video-Conferencing System Evan Broder and C. Christoher Post Introductory Digital Systems Laboratory November 2, 2007 Abstract The goal of this project is to create a video/audio conferencing system. Video

### Students will understand 1. use numerical bases and the laws of exponents

Grade 8 Expressions and Equations Essential Questions: 1. How do you use patterns to understand mathematics and model situations? 2. What is algebra? 3. How are the horizontal and vertical axes related?

### Current Standard: Mathematical Concepts and Applications Shape, Space, and Measurement- Primary

Shape, Space, and Measurement- Primary A student shall apply concepts of shape, space, and measurement to solve problems involving two- and three-dimensional shapes by demonstrating an understanding of:

### Linear Filtering Part II

Linear Filtering Part II Selim Aksoy Department of Computer Engineering Bilkent University saksoy@cs.bilkent.edu.tr Fourier theory Jean Baptiste Joseph Fourier had a crazy idea: Any periodic function can

### Correlation and Convolution Class Notes for CMSC 426, Fall 2005 David Jacobs

Correlation and Convolution Class otes for CMSC 46, Fall 5 David Jacobs Introduction Correlation and Convolution are basic operations that we will perform to extract information from images. They are in

### Characteristics and statistics of digital remote sensing imagery

Characteristics and statistics of digital remote sensing imagery There are two fundamental ways to obtain digital imagery: Acquire remotely sensed imagery in an analog format (often referred to as hard-copy)

### Investigation of Color Aliasing of High Spatial Frequencies and Edges for Bayer-Pattern Sensors and Foveon X3 Direct Image Sensors

Investigation of Color Aliasing of High Spatial Frequencies and Edges for Bayer-Pattern Sensors and Foveon X3 Direct Image Sensors Rudolph J. Guttosch Foveon, Inc. Santa Clara, CA Abstract The reproduction

### T = 1 f. Phase. Measure of relative position in time within a single period of a signal For a periodic signal f(t), phase is fractional part t p

Data Transmission Concepts and terminology Transmission terminology Transmission from transmitter to receiver goes over some transmission medium using electromagnetic waves Guided media. Waves are guided

### Auto-Tuning Using Fourier Coefficients

Auto-Tuning Using Fourier Coefficients Math 56 Tom Whalen May 20, 2013 The Fourier transform is an integral part of signal processing of any kind. To be able to analyze an input signal as a superposition

### Original Lecture Notes developed by

Introduction to ADSL Modems Original Lecture Notes developed by Prof. Brian L. Evans Dept. of Electrical and Comp. Eng. The University of Texas at Austin http://signal.ece.utexas.edu Outline Broadband

### Application Note 9. Digital FIR Decimator & Analog Lowpass

Application Note 9 App Note Application Note 9 Highlights Multirate FIR Design Cascade Analog Lowpass Circuit Optimization Comb Filter Correction Sin(x)/x Correction n Design Objective 16:1 FIR Decimation

### CM0340 SOLNS. Do not turn this page over until instructed to do so by the Senior Invigilator.

CARDIFF UNIVERSITY EXAMINATION PAPER Academic Year: 2008/2009 Examination Period: Examination Paper Number: Examination Paper Title: SOLUTIONS Duration: Autumn CM0340 SOLNS Multimedia 2 hours Do not turn

### Using Microcontrollers in Digital Signal Processing Applications

Using Microcontrollers in Digital Signal Processing Applications 1. Introduction Digital signal processing algorithms are powerful tools that provide algorithmic solutions to common problems. For example,

### (Refer Slide Time: 06:10)

Computer Graphics Prof. Sukhendu Das Dept. of Computer Science and Engineering Indian Institute of Technology, Madras Lecture - 43 Digital Image Processing Welcome back to the last part of the lecture

### Spectrum Level and Band Level

Spectrum Level and Band Level ntensity, ntensity Level, and ntensity Spectrum Level As a review, earlier we talked about the intensity of a sound wave. We related the intensity of a sound wave to the acoustic

### V.D.U. / Monitor glossary pg. 153. Display Screen vs. Monitor. Types of Monitors. 1. Cathode Ray Tube (CRT)

V.D.U. / Monitor glossary pg. 153 A display device is an output device that conveys text, graphics, and video information to the user. Information on a display device is called a soft copy because it exists

### Labview VI Example Virtual Filters Written by: Dan Lankow 2014

PH-315 Portland State University Labview VI Example Virtual Filters Written by: Dan Lankow 2014 1. ABSTRAT For this lab, you will be introduced to Labview. You will be implementing a Low Pass, High Pass,

### The Calculation of G rms

The Calculation of G rms QualMark Corp. Neill Doertenbach The metric of G rms is typically used to specify and compare the energy in repetitive shock vibration systems. However, the method of arriving

### Vocabulary Words and Definitions for Algebra

Name: Period: Vocabulary Words and s for Algebra Absolute Value Additive Inverse Algebraic Expression Ascending Order Associative Property Axis of Symmetry Base Binomial Coefficient Combine Like Terms

### 8-1 Adding and Subtracting Polynomials

Determine whether each expression is a polynomial. If it is a polynomial, find the degree and determine whether it is a monomial, binomial, or trinomial. 1. 7ab + 6b 2 2a 3 yes; 3; trinomial 2. 2y 5 +

### Data Storage 3.1. Foundations of Computer Science Cengage Learning

3 Data Storage 3.1 Foundations of Computer Science Cengage Learning Objectives After studying this chapter, the student should be able to: List five different data types used in a computer. Describe how

### Grafisk produktion och Tryckkvalitet

Grafisk produktion och Tryckkvalitet Rastrering och objektiva kvalitetsmått Sasan Gooran (VT 2007) Why halftoning? Original Halftoned Prepress Halftoning Print Digital Image Image Conventional Halftoning

### Conceptual Framework Strategies for Image Compression: A Review

International Journal of Computer Sciences and Engineering Open Access Review Paper Volume-4, Special Issue-1 E-ISSN: 2347-2693 Conceptual Framework Strategies for Image Compression: A Review Sumanta Lal

### JPEG Image Compression by Using DCT

International Journal of Computer Sciences and Engineering Open Access Research Paper Volume-4, Issue-4 E-ISSN: 2347-2693 JPEG Image Compression by Using DCT Sarika P. Bagal 1* and Vishal B. Raskar 2 1*

### CS425 Lab: Frequency Domain Processing

CS425 Lab: Frequency Domain Processing 1. Discrete Fourier Transform See section 14.1 in your textbook This is a brief review of the Fourier transform. An in-depth discussion of the Fourier transform is

### Frequency Response of FIR Filters

Frequency Response of FIR Filters Chapter 6 This chapter continues the study of FIR filters from Chapter 5, but the emphasis is frequency response, which relates to how the filter responds to an input

### This unit will lay the groundwork for later units where the students will extend this knowledge to quadratic and exponential functions.

Algebra I Overview View unit yearlong overview here Many of the concepts presented in Algebra I are progressions of concepts that were introduced in grades 6 through 8. The content presented in this course