You are on page 1of 64

Practical-1 Introduction to image processing toolbox

Introduction.
WHAT IS THE IMAGE PROCESSING TOOLBOX? The image processing Toolbox is a collection of functions that extend the capability of MATLAB. The toolbox support wide range of image processing, including • • • • • • • • • Spatial image transformation Morphological operations Neighbourhood and block operations Linear filtering and filter design Transforms Image analysis and enhancement Image registration Deblurring Region of interest operations

Many of the toolbox functions are matlab M-files. A Series of MATLAB statements that implement specialised image processing algorithms. You can view matlab codes for these functions using the statement type function _name.

Related Products The Math Works provides several product products that are especially relevant to the kinds of tasks you can perform with image processing toolbox. For more information about any of these products see either • The online documentations for that product if it is installed or if you are reading the documentation on CD. The Math works Web site at htto://www.mathworks.com; see the products section. The toolboxes listed below all include function that extends MATLAB. The block sets all include block that extend simulink.

Product Dsp Blockset Image acquisition toolbox

Description Design and simulate dsp system Connect to image acquisition hardware and bring frames of image data into matlab workspace Analyze and visualize geographically based information Perform signal processing algorithm development ,analysis and

Mapping toolbox Signal processing toolbox Wavelet toolbox

Analyze, compress and denoise signals and images using wavelet techniques

%it display image cameraman.Practical -2 DIGITAL IMAGE FUNDAMENTALS Objectives: • • Read and display image. %it reads the image and store it in a variable i. .tif').tif. Write the image >> i=imread('cameraman.png').'cameraman. >> imshow(i). OUTPUT: OUTPUT: >> imwrite(i. %writes the image to file & you can specify othe format also.

OUTPUT: >> imfinfo('cameraman.png') %return information about image. ans = Filename: 'cameraman.png' FileModDate: '19-Jul-2009 13:58:47' FileSize: 38267 Format: 'png' FormatVersion: [] Width: 256 Height: 256 BitDepth: 8 ColorType: 'grayscale' FormatSignature: [137 80 78 71 13 10 26 10] Colormap: [] .

Histogram: [] InterlaceType: 'none' Transparency: 'none' SimpleTransparencyData: [] BackgroundColor: [] RenderingIntent: [] Chromaticities: [] Gamma: [] XResolution: [] YResolution: [] ResolutionUnit: [] XOffset: [] YOffset: [] OffsetUnit: [] SignificantBits: [] ImageModTime: '19 Jul 2009 08:28:47 +0000' Title: [] Author: [] Description: [] Copyright: [] CreationTime: [] Software: [] Disclaimer: [] Warning: [] Source: [] .

imhist(a). subplot(1.1).Comment: [] OtherText: [] %To read image. subplot(1. to display image and its histogram in same plot Solution: a=imread('cameraman.2). OUTPUT: . imshow(a).tif').2.2.

% To display image in separate figure.. a=imread('moon.tif'); % To display each image in a separate figure figure(1); imshow(a); b=imread('tire.tif'); figure(2); imshow(b); c=imread('trees.tif'); figure(3); imshow(c);

OUTPUT:

% To display image in same figure... a=imread('moon.tif'); %To display multiple images in the same figure subplot(1,3,1); imshow(a); b=imread('tire.tif'); subplot(1,3,2); imshow(b); c=imread('trees.tif'); subplot(1,3,3); imshow(c); OUTPUT:

Objectives...

→ Image Sampling → Image Cropping. → Flipping of image.. → Rotate image by 30®. → Adjust contrast using histogram equalization... → Create negative image. →Adjusting contrasts.... Solution: a=imread('cameraman.tif'); subplot(2,2,1);

2).tif'). OUTPUT: % Image Cropping. imshow(c). OUTPUT: . imcrop(a). imhist(c). a=imread('moon.3). subplot(2.4).2. imhist(a).2. subplot(2.2. %it will plot histogram c=a(1:2:a(1).b=imshow(a). %To crop image by selecting interested area using mouse.1:2:a(2)). subplot(2.

3).2. subplot(2.. imshow(a). . imshow(c).2).2. b=size(a).4). %it will show size of image c=a(b(1):-1:1.2. subplot(2.b(2):-1:1).tif'). imhist(a).. subplot(2.1). a=imread('cameraman.%Cropping the image % Flipping of image. imhist(c).2. %Flip image by 180 subplot(2.

subplot(2.tif'). %Rotating the image by 35 counter clockwise subplot(2.2. imshow(b). imshow(a). . subplot(2.2. a=imread('cameraman.2.3). imhist(a).OUTPUT: %Flip the image OUTPUT: % Rotate image by 30®.2). b=imrotate(a.35).1).

2. imshow(b). subplot(2.2. subplot(2. .2..2).2. subplot(2.tif').subplot(2. imhist(a). imshow(a).3). %It shows the graph and contrast also. OUTPUT: % Adjust contrast using histogram equalization. a=imread('pout. imhist(b).4).. subplot(2.4).1).2. b=histeq(a).

2. subplot(2.[0 1]. imhist(a). .4).2). imhist(b).2. subplot(2.[1 0]).tif'). subplot(2. b=imadjust(a.2. subplot(2.2. %It shows negative image of original image.imhist(b).3). OUTPUT: % Create negative image. imshow(b). imshow(a). a=imread('cameraman.1)..

[0 1].[0 0.3]. imshow(i). . imhist(i).[].2.5].2.2 0..tif'). i=imread('tire.2 0.[0.[1 0]). d=imadjust(i..3 0.5]).[0 1]).3)..2.7]). b=imadjust(i. subplot(5. subplot(5. c=imadjust(i.[0.OUTPUT: %Adjusting contrasts.[0.2). subplot(5. a=imadjust(i.1).

2. imshow(b). imhist(d).2. imshow(d). subplot(5. subplot(5. subplot(5.9).2.2. . subplot(5. subplot(5.4). subplot(5.6).2. OUTPUT.7). subplot(5. imhist(a). imhist(c).10).2. imhist(b).8).2. imshow(c).5).imshow(a).

........................................................50 0 0 0 5 0 10 0 10 5 20 0 20 5 10 00 50 0 0 0 50 00 5 0 10 0 10 5 20 0 20 5 0 0 5 0 10 0 10 5 20 0 20 5 10 00 0 0 5 0 10 0 10 5 20 0 20 5 50 0 0 0 5 0 10 0 10 5 20 0 20 5 Conclusion: .............................................................................................................................................................................................................................................................................................................................................................................................. ..................................... Practical-3 ............................................... ........... ............................................ ......................................................................................................................................................................................................................................

title('Histogram of the original image'). subplot(2. imshow(I).4). plot(L). % Obtaining the histogram of the adjusted image L=imhist(K). title('Original image'). % Increasing the contrast of the adjusted image K=imadjust(I).2. % Obtaining the histogram of the original image J=imhist(I).2).3). plot(J). OUTPUT: . subplot(2. figure(1).2. imshow(K) title('Intensity Adjusted image').1). Solution:Read and display the original image I=imread('trees. 1) Program to increase the contrast in a low contrast grayscale image and display the histogram of the output image. Compare the histogram before and after intensity adjustments.tif'). subplot(2. title('Histogram of the Adjusted image'). subplot(2.2.2.Image Enhancement In The Spatial Domain.

title('Image before gamma correction'). imshow(b).2).0. imshow(a).5).3. subplot(1.tif').3. figure(1). K=imabsdiff(a.[].1).3). imshow(b). imshow(X).[]. title('Image after gamma correction').5. figure(2). subplot(1.3.b). subplot(1.2] Program to display an image with and without gamma correction →Find absolute difference of two images X=imread('forest. subplot(1.1). % Gamma correction is applied & the value of gamma is taken as 0. b=imadjust(a.3. subplot(1. .3. title('Image after gamma correction').3).

imshow(K. title('Absolute difference between two images').[]). Output: In n ityfo a img te s rmt a e I a ea rg m ac r e tio mg fte a m or c n .

.2).I a eb f reg m ac re tio mg eo a m or c n I a ea rg m ac re tio mg fte a m or c n A d no tw img s d itio f o a e 3] Program to adjust intensity values using:a) Histogram Equalization b) Contrast-Limited Adaptive Histogram Equalization Solution:I=imread('cameraman.tif').1).2. title('Histogram of the original image'). imshow(I). plot(J). subplot(3. J=imhist(I). title('Original image').3). subplot(3. % Adjusting intensity values using Histogram equalization K=histeq(I).2.2. imshow(K). subplot(3.

title('Histogram of the Equalized Image').2.2.5). % Adjusting intensity values using contrast-Limited Adaptive Histogram Equalization M=adapthisteq(I). title('Histogram of the Contrast-Limited adaptive Equalized Image').6). L=imhist(K). plot(N). N=imhist(M).4). OUTPUT: . imshow(M). title('Contrast.Limited Adaptive Histogram Equalized image'). subplot(3. subplot(3.2. subplot(3. plot(L).title('Histogram equalized image').

2. imshow(i). subplot(2. . subplot(2.2.4). imhist(b).2.2.3). subplot(2.1).2). title('histogram of histequalization. imhist(i). title('histogram of original').'). subplot(2. b=histeq(i).• Enhance contrast of an intensity image using histogram equalization. i=imread('cameraman. imshow(b). Display the histogram of the original image and compare it with histogram of the processed image.tif').

........................................................................................................... Practical-4 Filtering in spatial domain ...........................Conclusion: ........ .................................................................................... .................................................................................................................................................................................................................................................................................................................................................................................................................. ............................ .......................................................................................................................................................

imshow(a). imshow(result).tif').j)/9 + a(i+1. title('original image').j+1)/9 + a(i.1.j+1)/9). end end subplot(2. result(i.j-1)/9 + a(i.• Create 3x3 mask in m file without using its function and implement on image.1).j+1)/9 + a(i+1.j-1)/9 + a(i-1. . subplot(2. %creating 3x3 mask a=imread('tire. for j=2:(l-1). [k l]=size(a).1.j)/9 + a(i-1. for i=2:(k-1).j)=(a(i-1.j-1)/9 + a(i+1. title('smoothed image using average filter of 3x3').j)/9 + a(i.2).

subplot(3. title('difference between original and the filtered 8'). figure(). result3=imfilter(i.result2).tif').35). b=imhist(result1). result1=imfilter(i. result2=imfilter(i. . title('smoothing with mask of 64x64'). h=fspecial('average'.• Program to illustrate the effect of smoothing filter as a function of filter size. subplot(3. title('smoothing with mask of 16x16').16. title('smoothing with mask of 35x35'). subplot(1.2.16).'replicate').2.8).2).h. h=fspecial('average'.h. 16. plot(b).5). 64. %absolute difference between original and using mask of 8.'replicate').2. subplot(3.2. imshow(result4).2.2.4).h.64).1).2).3). h=fspecial('average'. figure(). subplot(1. 128. title('smoothing with mask of 8x8'). imshow(i). Solution: i=imread('cameraman. imshow(result2).result1).'replicate'). Plot difference between the original and filtered image and their corrosponding histogram.2. title('histogram of filter size 8'). result4=imfilter(i. imshow(result1).35. subplot(3.h. %smoothing mask of 8x8 h=fspecial('average'. figure(). a=imabsdiff(i.64 a=imabsdiff(i. Obtain the result of filtering with mask of 8. title('original image').1).'replicate'). imshow(a). subplot(3. imshow(result3).

b=imhist(result4).1).2). imshow(a).2. plot(b). subplot(1.result3).subplot(1. subplot(1. imshow(a). a=imabsdiff(i. title('histogram of filter size 35').1). plot(b).result4). figure(). subplot(1. plot(b). figure(). b=imhist(result2). title('histogram of filter size 64').2).2. subplot(1. subplot(1.1). imshow(a). title('difference between original and the filtered 35').2.2). title('histogram of filter size 16').2. title('difference between original and the filtered 64').2. a=imabsdiff(i. title('difference between original and the filtered 16').2. b=imhist(result3). .

.

.

.

.

15. imshow(motion). Gaussianed image. %blurred image h=fspecial('disk'.h.45).2). Sobeled image. imshow(i). Sharpened image. . %obtaining motion blurred image h=fspecial('motion'. subplot(3.15). Blurred image.1).• Program to illustrate the grayscale image and to obtain the following types of filtered images. motion=imfilter(i. title('original image'). subplot(3. Avereged image.3.tif'). title('image formed due to motion').3. Solution: i=imread('cameraman.Motion blurred image.'replicate'). Laplaced image.

8). title('image formed due to gaussian').3. imshow(average). gaussian=imfilter(i. subplot(3.h.h.5). imshow(gaussian). title('image formed due to sharpining'). imshow(sobel). title('image formed due to laplaced').3.h.[3 3].'replicate'). average=imfilter(i.h. laplacian=imfilter(i.h. title('image formed due to disk').4).3).3). .3). subplot(3.'replicate'). imshow(laplacian). sharp=imfilter(i.disk=imfilter(i. subplot(3.6). subplot(3.'replicate'). title('image formed due to averaging').3. imshow(disk). %laplaced image h=fspecial('laplacian'.7).h. sobel=imfilter(i.0. subplot(3. title('image formed due to sobel').3. subplot(3. %sharpened image h=fspecial('unsharp'). %gaussianed image h=fspecial('gaussian'.'replicate').'replicate'). %sobeled image h=fspecial('sobel').4). %averaged image h=fspecial('average'.3.0. imshow(sharp).3.'replicate').

0. subplot(1. subplot(1. title('original image').02).2. Solution: i=imread('cameraman.2). imshow(i). imshow(j). %filtering the image with 3x3 averaging filter .2. title('salt and pepper noise'). %adding the salt and pepper noise to original image j=imnoise(i. figure().1).• Program to illustrate median filtering over averaging filter in sitution when image is corrupted with salt and pepper noise.'salt & pepper'.tif').

title('averaged image with filter size of 3x3'). . title('difference between original image and median filtered image').3).'symmetric'). %absolute difference between the original and averaged image x=imabsdiff(i.1). median=medfilt2(i.1). imshow(y). figure(). imshow(x).h.average). title('median image with mask of 3x3'). average=imfilter(i. y=imabsdiff(i. title('difference between original median image').median).2). subplot(1.'symmetric').2). subplot(1. subplot(1.2.2. figure().[3 3].h=fspecial('average'. subplot(1.2. imshow(average). imshow(median).2.

.

.

..................................................... ........................................................................................................ ................................................................................................................................................................................................................................... .......................................................................................................................................................................Conclusion: ....... ......................................................................................................................................................... ................................................................................................

n).v) = sqrt((u-M/2)^2+(v-N/2). Gaussian low pass filter] function [H. for u=1:M for v=1:N D(u.^2). end end D=fftshift(D). . switch type case 'Ideal' H = double(D<=D0).D0.M. n = 1.PRACTICAL – 5 IMAGE ENHANCEMENT IN THE FREQUENCY DOMAIN %Low pass filter function[Ideal. case 'Butterworth' if nargin == 4.D] = lpfilter(type.N. butterworth.

filt = lpfilter('Ideal'.2.30. Also obtain the corresponding frequency spectrum and filter displayed as an image. . imshow(lf_ip. lf_ip=fftshift(f_ip). imshow(ip) title('Original Image') figure./(2*(D0^2))). • Ideal low pass filter • Gaussian low pass filter • Butterworth low pass filter %Ideal Low pass filtered image with cutoff frequecy set at radius value of 5 clc clear all ip = imread('peppers. f_op = filt. f_ip = fft2(ip).N.*f_ip. lf_op=log(1+abs(f_op)).80 resp./D0).[]) title('Frequency Spectrum') subplot(1./(1 + (D. ip=double(ip). filt=fftshift(filt). lf_ip = log(1 + abs(lf_ip)).2.15. case 'Gaussian' H = exp(-(D.png').1).^2).N] = size(ip). subplot(1. op = real(ifft2(f_op)).1).5). ip=uint8(ip).2. subplot(1.^(2*n)). otherwise error('Unknown filter type.') end 1] Program to illustrate the results of following low pass filters with cut of frequencies set at radii values of 5. ip=rgb2gray(ip).2).end H = 1. [M.M.

*f_ip.2. f_op = filt.[]). filt=fftshift(filt).2. title('Filter Frequency domain') subplot(1. imshow(filt.N. title('Ideal Lowpass filtered image [cutoff freq=30]') %Ideal Low pass filtered image with cutoff frequecy set at radius value of 80 figure. filt=fftshift(filt). subplot(1.15). title('Ideal Lowpass filtered image [cutoff freq=15]') %Ideal Low pass filtered image with cutoff frequecy set at radius value of 30 figure.[]). .M.[]).N.2.30).[]).2). op = real(ifft2(f_op)). title('Filter Frequency domain') subplot(1. filt = lpfilter('Ideal'. imshow(op. imshow(op.2.*f_ip. filt=fftshift(filt). imshow(op.2). lf_op=log(1+abs(f_op)). imshow(filt. filt = lpfilter('Ideal'. subplot(1. lf_op=log(1+abs(f_op)).imshow(filt.1).2. title('Filter Frequency domain') subplot(1. title('Ideal Lowpass filtered image [cutoff freq=5]') %Ideal Low pass filtered image with cutoff frequecy set at radius value of 15 figure.*f_ip.[]). subplot(1.M. lf_op=log(1+abs(f_op)).1).M.N. op = real(ifft2(f_op)). f_op = filt.80).1). filt = lpfilter('Ideal'. f_op = filt.[]).2. op = real(ifft2(f_op)).2).

2.imshow(filt. filt = lpfilter('Ideal'.[]). imshow(op.2. lf_op=log(1+abs(f_op)).*f_ip. f_op = filt.M.[]). imshow(op.2). title('Ideal Lowpass filtered image [cutoff freq=230]') OUTPUT: . imshow(filt.N. filt=fftshift(filt). subplot(1. title('Filter Frequency domain') subplot(1.230).2). op = real(ifft2(f_op)).2. title('Filter Frequency domain') subplot(1. title('Ideal Lowpass filtered image [cutoff freq=80]') %Low pass filtered image with cutoff frequecy set at radius value of 230 figure.[]).1).[]).

Fe u n ySe tr m r qec pc u O in l Img r a ae ig .

F rFe u n yd min ilte r q e c o a Id a L wa sfiltee img [c to fe = ] el o ps r d a e u ff r q 5 F rFe u n yd min ilte r q e c o a I e l L wa sf r dimg [c tof fr q 1 ] d a o p s iltee a e u f e = 5 .

Filter Frequency domain Ideal Lowpass filtered image [cutoff freq=30] .

F rFe u n yd min ilte r q e c o a I e l L wa sf r dimg [c tof fr q 8 ] d a o p s iltee a e u f e = 0 .

f_ip = fft2(ip). ip=uint8(ip).png'). [M. op = real(ifft2(f_op)).*f_ip. lf_ip = log(1 + abs(lf_ip)).M. lf_ip=fftshift(f_ip).2).5).N. ip=rgb2gray(ip).1). f_op = filt. imshow(ip) title('Original Image') figure. ip=double(ip).[]) title('Frequency Spectrum') subplot(1. filt = lpfilter('Gaussian'. subplot(1. imshow(lf_ip.2.2.N] = size(ip). .F rFeunydmin ilte r qec o a I el Lwasf r dimg [c to fe=3] da o ps iltee ae u f r q 20 f %Gaussian Low pass filtered image with cutoff frequecy set at radius value of 5 clc clear all ip = imread('peppers.

title('Gaussian Lowpass filtered image [cutoff freq=15]') %Gaussian Low pass filtered image with cutoff frequecy set at radius value of 30 figure. .N.[]). imshow(filt.1).2. op = real(ifft2(f_op)).*f_ip.*f_ip.2. title('Gaussian Lowpass filtered image [cutoff freq=30]') %Gaussian Low pass filtered image with cutoff frequecy set at radius value of 80 figure. lf_op=log(1+abs(f_op)).2). op = real(ifft2(f_op)).M.M. title('Filter Frequency domain') subplot(1.2. lf_op=log(1+abs(f_op)). filt = lpfilter('Gaussian'. op = real(ifft2(f_op)).1).2.[]).2.2). title('Filter Frequency domain') subplot(1.[]).[]).15).2. filt = lpfilter('Gaussian'. imshow(filt.2). title('Filter Frequency domain') subplot(1.80). filt=fftshift(filt).[]).N.*f_ip. filt = lpfilter('Gaussian'. lf_op=log(1+abs(f_op)). subplot(1. imshow(op. f_op = filt. filt=fftshift(filt). imshow(filt.1).[]). filt=fftshift(filt). subplot(1.N.M. subplot(1. lf_op=log(1+abs(f_op)). title('Gaussian Lowpass filtered image [cutoff freq=5]') %Gaussian Low pass filtered image with cutoff frequecy set at radius value of 15 figure. imshow(op.filt=fftshift(filt).30). f_op = filt. imshow(op. f_op = filt.

2).2.*f_ip. title('Gaussian Lowpass filtered image [cutoff freq=230]') . title('Filter Frequency domain') subplot(1. f_op = filt. filt = lpfilter('Gaussian'. title('Filter Frequency domain') subplot(1. op = real(ifft2(f_op)).1).subplot(1. imshow(filt. filt=fftshift(filt). imshow(op.2.[]). title('Gaussian Lowpass filtered image [cutoff freq=80]') %Guassian Low pass filtered image with cutoff frequecy set at radius value of 230 figure.[]).2.1).[]). lf_op=log(1+abs(f_op)).2).2.N. subplot(1. imshow(op. imshow(filt.M.[]).230).

Fe u n ySe t r q ec pc O in l I g r am ig .

F rFe d min ilte r q o a Gu s nL filtee im [c tof f e = ] a s ia p r d g u f r q 5 .

F rFe d min ilte r q o a Gu s nL filtee im [c to fe = 5 a s ia p r d g u ff r q 1 ] .

F rFe d min ilte r q o a Gu s nL filtee im [c to fe = 0 a s ia p r d g u ff r q 3 ] .

F rFe d min ilte r q o a Gu s nL filtee im [c to fe = 0 a s ia p r d g u ff r q 8 ] .

2). ip=uint8(ip).2. subplot(1.[]) title('Frequency Spectrum') subplot(1.2. imshow(ip) title('Original Image') figure.5). filt = lpfilter('Butterworth'.M.N] = size(ip). imshow(lf_ip. f_ip = fft2(ip).1). .png'). lf_ip=fftshift(f_ip). ip=double(ip). [M.N. lf_ip = log(1 + abs(lf_ip)).F rFe u n yd min ilte r q e c o a Gu s nL wa sf r dimg [c to fe = 3 ] a s ia o p s iltee a e u ff r q 2 0 %Butterworth Low pass filtered image with cutoff frequecy set at radius value of 5 clc clear all ip = imread('onion. ip=rgb2gray(ip).

imshow(filt.2.15). subplot(1.N. filt=fftshift(filt). imshow(op.2). op = real(ifft2(f_op)).[]). filt = lpfilter('Butterworth'. title('Filter Frequency domain') subplot(1.[]).M.2. title('Butterworth Lowpass filtered image [cutoff freq=30]') %Butterworth Low pass filtered image with cutoff frequecy set at radius value of 80 figure. filt=fftshift(filt).[]). imshow(filt. lf_op=log(1+abs(f_op)).2).1).2).f_op = filt. imshow(op. title('Butterworth Lowpass filtered image [cutoff freq=5]') %Butterworth Low pass filtered image with cutoff frequecy set at radius value of 15 figure. filt = lpfilter('Butterworth'.2.*f_ip. . f_op = filt.1). title('Filter Frequency domain') subplot(1.1). f_op = filt.M. op = real(ifft2(f_op)). imshow(op.M.2. f_op = filt.*f_ip. subplot(1. op = real(ifft2(f_op)). op = real(ifft2(f_op)).*f_ip.2. title('Butterworth Lowpass filtered image [cutoff freq=15]') %Butterworth Low pass filtered image with cutoff frequecy set at radius value of 30 figure. imshow(filt. filt=fftshift(filt).N. subplot(1.[]). title('Filter Frequency domain') subplot(1. lf_op=log(1+abs(f_op)).30). filt = lpfilter('Butterworth'.[]).80). lf_op=log(1+abs(f_op)).2.*f_ip.N.[]).

lf_op=log(1+abs(f_op)). imshow(filt.2. title('Butterworth Lowpass filtered image [cutoff freq=230]') . op = real(ifft2(f_op)).2). subplot(1. title('Filter Frequency domain') subplot(1.2). title('Butterworth Lowpass filtered image [cutoff freq=80]') %Guassian Low pass filtered image with cutoff frequecy set at radius value of 230 figure. lf_op=log(1+abs(f_op)). subplot(1.[]).[]).2.2. imshow(op. imshow(op.filt=fftshift(filt). f_op = filt. title('Filter Frequency domain') subplot(1. filt=fftshift(filt).N. imshow(filt.*f_ip.230).[]).2. filt = lpfilter('Butterworth'.1).M.1).[]).

Fe u n yS e tr m r qec pc u O in l I a e r a mg ig .

F rFe u n yd min ilte r q e c o a Bttewr L wa sf r dimg [c toff e = ] u r oth o ps iltee a e u f r q 5 .

F rFe u n yd min ilte r q e c o a B ttewr L wa sfiltee img [c to f q 1 ] u r oth o p s r d a e u ff re = 5 .

F rFe u n yd min ilte r q e c o a Bttewr L wa sfiltee img [c to f e= 0 u r oth o p s r d a e u ff r q 3 ] .

F rFe u n yd min ilte r q e c o a B ttewr L wa sfiltee img [c to f q 8 ] u r oth o p s r d a e u ff re = 0 .

.......................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................... ................................................ .......... .................................................................................................................................. .....F rFe u n yd min ilte r q e c o a Bttewr L wa sf r dimg [c tof fr q 2 0 u r oth o p s iltee a e u f e = 3 ] \Conclusion: ............ ...............