TUGAS PENGOLAHAN CITRA

Disusun Oleh:
Nama : Yeni Laraswati (09011181520029)
Jurusan :Sistem Komputer
Dosen ​ rwin, M.Si
:E

JURUSAN SISTEM KOMPUTER

FAKULTAS ILMU KOMPUTER
UNIVERSITAS SRIWIJAYA
2018
1. 4.4 Tutorial 4.2: Basic Image Manipulation / 74
1. Load the image coins.png by executing the following statement:
I = imread(’coins.png’);

2. Load the image trees.tif.

[X,map] = imread(’trees.tif’);
3. Convert the indexed image X with color map map to an RGB image, X_rgb.
X_rgb = ind2rgb(X,map);

4. Convert the indexed image X with color map map to an intensity image.
X_gray = ind2gray(X,map);
5. We can verify that the new intensity image consists of pixel values in the range
[0, 255].
max(X_gray(:))
min(X_gray(:))

6. Convert the variable X_gray to class double.

X_gray_dbl = im2double(X_gray);
7. Use the imshow function (with the impixelinfo option) to display the
coins.png image that is currently loaded in the variable I.
imshow(I), impixelinfo

8. Display the indexed image trees.tif. The image data are stored in variable X and the color map
in map. Note that the impixelinfo option provides a clear hint that this is an indexed color image.
 imshow(X,map), impixelinfo

9. Use theimpixelfunction to explore interactively the pixel contents of selected points in the
image. Use the mouse to click on the points of interest: normal button clicks are used to select
pixels, pressing ​Backspace ​or ​Delete ​removes the previously selected pixel, a double-click adds
a final pixel and ends the selection, and pressing ​Return ​finishes the selection without adding a
final pixel.
RGB = imread(’peppers.png’);
[c,r,p] = impixel(RGB);
10. Use the improfile function to explore the contents of a line in the coins.png image that is
currently loaded in the variable I.
r1 = 17; c1 = 18; r2 = 201; c2 = 286;
imshow(I)
line([c1, c2], [r1, r2], ’Color’, ’g’, ’LineWidth’, 2);
figure
improfile(I, [c1, c2], [r1, r2]);
ylabel(’Gray level’);

11. Use the imtool function to display the image currently loaded in the variable X_rgb. Note
that a secondary window (Overview) will open as well. Explore the additional functionality,
including the possibility of measuring distances between two points within the image.
imtool(X_rgb)

12. Close any open figures (close all).

13. Execute the following statements to create a subplot with two images:
A = imread(’pout.tif’);
B = imread(’cameraman.tif’);
figure
subplot(1,2,1), imshow(A)
subplot(1,2,2), imshow(B)

14. Close any open figures.

15. Display the coins.png (loaded in variable I) and the trees.tif (loaded in variable X and its
color map in variable map) images in a subplot. Execute each statement at a time to see the effect
on the images as they are displayed.
figure
subplot(1,2,1), imshow(I)
subplot(1,2,2), imshow(X,map)

16. Use the subimage function to display multiple images with different color maps.
figure
subplot(1,2,1), subimage(I), axis off
subplot(1,2,2), subimage(X,map), axis off

17. Manually convert the intensity image coins (loaded in the variable I) to an indexed image and
then to RGB. Note that the trees image (loaded in variable X with its color map in variable map)
has already been converted to RGB in
step 3 (saved in variable X_rgb).
[I_ind,I_map] = gray2ind(I,256);
I_rgb = ind2rgb(I_ind,I_map);
18. Display the truecolor images using the imshow function.
figure
subplot(1,2,1), imshow(I_rgb)
subplot(1,2,2), imshow(X_rgb)

19. Use imwrite to save two of the modified images in this tutorial to files for further use. Use
the JPEG format for one of them and the PNG extension for
the other. For example,
imwrite(X_rgb, ’rgb_trees.jpg’);
imwrite(X_gray, ’gray_trees.png’);
2. 6.3 Tutorial 6.1: Arithmetic Operations / 113
1. Use the imadd function to brighten an image by adding a constant (scalar)
value to all its pixel values.
I = imread(’tire.tif’);
I2 = imadd(I,75);
figure
subplot(1,2,1), imshow(I), title(’Original Image’);
subplot(1,2,2), imshow(I2), title(’Brighter Image’);
2. Use the imadd function to blend two images.
Ia = imread(’rice.png’);
Ib = imread(’cameraman.tif’);
Ic = imadd(Ia,Ib);
figure
imshow(Ic);

3. Close all open figures and clear all workspace variables.

4. Load two images and display them.
I = imread(’cameraman.tif’);
J = imread(’cameraman2.tif’);
figure
subplot(1,2,1), imshow(I), title(’Original Image’);
 subplot(1,2,2), imshow(J), title(’Altered Image’);

5. Subtract both images and display the result.

diffim = imsubtract(I,J);
figure
subplot(2,2,1), imshow(diffim), title(’Subtracted Image’);

6. Use the zoom tool to zoom into the right area of the difference image about
halfway down the image. You will notice that a small region of pixels is faintly
white.
7. To zoom back out, double-click anywhere on the image.
8. Calculate the absolute difference. Make sure Figure 2 is selected before executing
this code.
diffim2 = imabsdiff(I,J);
subplot(2,2,2), imshow(diffim2), title(’Abs Diff Image’);

9. Use the zoom-in tool to inspect the new difference image

10. Show scaled versions of both difference images.
subplot(2,2,3), imshow(diffim,[]), ...
title(’Subtracted Image Scaled’);
subplot(2,2,4), imshow(diffim2,[]), ...
title(’Abs Diff Image Scaled’);

11. Use the zoom tool to see the differences between all four difference images
12. Close all open figures and clear all workspace variables.
13. Use immultiply to dynamically scale the moon image.
I = imread(’moon.tif’);
I2 = imadd(I,50);
I3 = immultiply(I,1.2);
 figure
subplot(1,3,1), imshow(I), title(’Original Image’);
subplot(1,3,2), imshow(I2), title(’Normal Brightening’);
subplot(1,3,3), imshow(I3), title(’Dynamic Scaling’);

14. Close all open figures and clear all workspace variables.
15. Create an artificial 3D planet by using the immultiply function to multiply
the earth1 and earth2 images.
I = im2double(imread(’earth1.tif’));
J = im2double(imread(’earth2.tif’));
K = immultiply(I,J);
figure
subplot(1,3,1), imshow(I), title(’Planet Image’);
(1,3,2), imshow(J), title(’Gradient’);
subplot(1,3,3), imshow(K,[]), title(’3D Planet’);

16. Close all open figures and clear all workspace variables.
17. Use image division to dynamically darken the moon image.
I = imread(’moon.tif’);
 I2 = imdivide(I,2);
Figure
subplot(1,3,1), imshow(I), title(’Original Image’);
subplot(1,3,2), imshow(I2), title(’Darker Image w/ Division’)

18. Display the equivalent darker image using image multiplication.

I3 = immultiply(I,0.5);
subplot(1,3,3), imshow(I3), ...
title(’Darker Image w/ Multiplication’);

19. Close all open figures and clear all workspace variables.

20. Load the images that will be used for background subtraction.
 notext = imread(’gradient.tif’);
text = imread(’gradient_with_text.tif’);
figure, imshow(text), title(’Original Image’);

21. Show how thresholding fails in this case.

level = graythresh(text);
BW = im2bw(text,level);
figure, imshow(BW)
22. Divide the background from the image to get rid of the background.
fixed = imdivide(text,notext);
figure
subplot(1,3,1), imshow(text), title(’Original Image’);
subplot(1,3,2), imshow(notext), title(’Background Only’);
subplot(1,3,3), imshow(fixed,[]), title(’Divided Image’)

6.4 Tutorial 6.2: Logic Operations and Region of Interest Processing

1. Use theMATLAB help system to learn howto use the roipoly function when
only an image is supplied as a parameter.
2. Use the roipoly function to generate a mask for the pout image.
I = imread(’pout.tif’);
bw = roipoly(I);
3. Convert the mask image to class uint8.
bw2 = uint8(bw);

4. Use the bitand function to compute the logic AND between the original
image and the new mask image.
I2 = bitand(I,bw2);
imshow(I2);

5. Use the bitcmp function to generate a complemented version of the bw2 mask.
bw_cmp = bitcmp(bw2);
figure
subplot(1,2,1), imshow(bw2), title(’Original Mask’);
subplot(1,2,2), imshow(bw_cmp), title(’Complemented Mask’);
6. Use bitor to compute the logic OR between the original image and the complemented
mask.
I3 = bitor(I,bw_cmp);
figure, imshow(I3)

7. Complement an image using the imcomplement function.

bw_cmp2 = imcomplement(bw2);
8. Close all open figures and clear all workspace variables.
9. Use the bitxor function to find the difference between two images.
I = imread(’cameraman.tif’);
I2 = imread(’cameraman2.tif’);
I_xor = bitxor(I,I2);
figure
subplot(1,3,1), imshow(I), title(’Image 1’);
subplot(1,3,2), imshow(I2), title(’Image 2’);
subplot(1,3,3), imshow(I_xor,[]), title(’XOR Image’);

10. Close all open figures and clear all workspace variables.
11. Read in image and calculate an adjusted image that is darker using the imdivide function.
 I = imread(’lindsay.tif’);
I_adj = imdivide(I,1.5);

12. Generate a mask by creating a region of interest polygon.

bw = im2uint8(roipoly(I));

13. Use logic operators to show the darker image only within the region of interest, while
displaying the original image elsewhere.
bw_cmp = bitcmp(bw); %mask complement
roi = bitor(I_adj,bw_cmp); %roi image
not_roi = bitor(I,bw); %non_roi image
new_img = bitand(roi,not_roi); %generate new image
imshow(new_img) %display new image
3. 7.6 Tutorial 7.1: Image Cropping, Resizing, Flipping,
and Rotation / 138
1. Open the cameraman image and use the ​Crop Image ​option in the Image Tool
(imtool) toolbar to crop it such that only the portion of the image containing
the tallest building in the background is selected to become the cropped image.
Pay attention to (and write down) the coordinates of the top left and bottom
right corners as you select the rectangular area to be cropped. You will need
this information for the next step.
2. Double-click inside the selected area to complete the cropping operation.
3. Save the resulting image using the ​File > Save as... ​option in the imtool
menu. Call it cropped_building.png.
I = imread(’cameraman.tif’);
imtool(I)

4. Open and display the cropped image.

I2 = imread(’cropped_building.png’);
imshow(I2)
5. We shall now use the coordinates recorded earlier to perform a similar cropping
from a script.
6. The imcrop function expects the crop rectangle—a four-element vector
[xmin ymin width height]—to be passed as a parameter.
7. Perform the steps below replacing my values for x1, y1, x2, and y2 with the
values you recorded earlier.
x1 = 186; x2 = 211; y1 = 105; y2 = 159;
xmin = x1; ymin = y1; width = x2-x1; height = y2-y1;
I3 = imcrop(I, [xmin ymin width height]);
imshow(I3)

8. Enlarge the cameraman image by a scale factor of 3. By default, the function

uses bicubic interpolation.
I_big1 = imresize(I,3);
figure, imshow(I), title(’Original Image’);
figure, imshow(I_big1), ...
title(’Enlarged Image w/ bicubic interpolation’);

9. Use the imtool function to inspect the resized image, I_big1.

imtool(I_big1)
10. Scale the image again using nearest-neighbor and bilinear interpolations.
I_big2 = imresize(I,3,’nearest’);
I_big3 = imresize(I,3,’bilinear’);
figure, imshow(I_big2), ...
title(’Resized w/ nearest-neighbor interpolation’);
figure, imshow(I_big3), ...
title(’Resized w/ bilinear interpolation’);

11. Close any open figures.

12. Reduce the size of the cameraman image by a factor of 0.5 in both dimensions.
I_rows = size(I,1);
I_cols = size(I,2);
I_sm1 = I(1:2:I_rows, 1:2:I_cols);
figure, imshow(I_sm1);
13. Shrink the image using the imresize function.
I_sm2 = imresize(I,0.5,’nearest’);
I_sm3 = imresize(I,0.5,’bilinear’);
I_sm4 = imresize(I,0.5,’bicubic’);
figure, subplot(1,3,1), imshow(I_sm2), ...
title(’Nearest-neighbor Interpolation’);
subplot(1,3,2), imshow(I_sm3), title(’Bilinear Interpolation’);
subplot(1,3,3), imshow(I_sm4), title(’Bicubic Interpolation’);

14. Close all open figures and clear all workspace variables.
15. Flip the cameraman image upside down.
16. Flip the cameraman image from left to right.
I = imread(’cameraman.tif’);
J = flipud(I);
K = fliplr(I);
subplot(1,3,1), imshow(I), title(’Original image’)
subplot(1,3,2), imshow(J), title(’Flipped upside-down’)
subplot(1,3,3), imshow(K), title(’Flipped left-right’)
17. Close all open figures and clear all workspace variables.
18. Rotate the eight image by an angle of 35◦.
I = imread(’eight.tif’);
I_rot = imrotate(I,35);
imshow(I_rot);

19. Rotate the same image using bilinear interpolation.

I_rot2 = imrotate(I,35,’bilinear’);
figure, imshow(I_rot2)
20. Rotate the same image, but this time crop the output.
I_rot3 = imrotate(I,35,’bilinear’,’crop’);
figure, imshow(I_rot3)

7.7 Tutorial 7.2: Spatial Transformations and Image Registration / 142

1. Open the cameraman image.
2. Use maketform to make an affine transformation that resizes the image by a
factor [​sx, sy]​ . The maketform function can accept transformation matrices
of various sizes for ​N-​ dimensional transformations. But since imtransform
only performs 2D transformations, you can only specify 3 × 3 transformation
matrices. For affine transformations, the first two columns of the 3 × 3 matrices
will have the values ​a​0​, a​1​, a​2​, b​0​, b​1​, b​2 from Table 7.1, whereas the last
column must contain0 0 1.
3. Use imtransform to apply the affine transformation to the image.
4. Compare the resulting image with the one you had obtained using imresize.
I1 = imread(’cameraman.tif’);
sx = 2; sy = 2;
T = maketform(’affine’,[sx 0 0; 0 sy 0; 0 0 1]’);
I2 = imtransform(I1,T);
imshow(I2), title(’Using affine transformation’)
I3 = imresize(I1, 2);
figure, imshow(I3), title(’Using image resizing’)

5. Use maketform to make an affine transformation that rotates an image by an

angle ​θ​.
6. Use imtransform to apply the affine transformation to the image.

7. Compare the resulting image with the one you had obtained using imrotate.
I1 = imread(’cameraman.tif’);
 theta = 35*pi/180
xform = [cos(theta) sin(theta) 0; -sin(theta) cos(theta) 0; 0 0 1]’
T = maketform(’affine’,xform);
I4 = imtransform(I1, T);
imshow(I4), title(’Using affine transformation’)
I5 = imrotate(I1, 35);
figure, imshow(I5), title(’Using image rotating’)

8. Use maketform to make an affine transformation that translates an image by

_x,_y​.
9. Use imtransform to apply the affine transformation to the image and use a
fill color (average gray in this case) to explicitly indicate the translation.
10. Display the resulting image.
I1 = imread(’cameraman.tif’);
delta_x = 50;
delta_y = 100;
xform = [1 0 delta_x; 0 1 delta_y; 0 0 1]’
tform_translate = maketform(’affine’,xform);
I6 = imtransform(I1, tform_translate,...
’XData’, [1 (size(I1,2)+xform(3,1))],...
’YData’, [1 (size(I1,1)+xform(3,2))],...
’FillValues’, 128 );
figure, imshow(I6)

11. Use maketform to make an affine transformation that performs shearing by

a factor [sh​x, ​sh​y​] on an input image.
12. Use imtransform to apply the affine transformation to the image.
13. Display the resulting image.
I = imread(’cameraman.tif’);
sh_x = 2; sh_y = 1.5;
xform = [1 sh_y 0; sh_x 1 0; 0 0 1]’
T = maketform(’affine’,xform);
I7 = imtransform(I1, T);
imshow(I7)

14. Open the base image (Figure 7.9a) and the unregistered image (Figure 7.9b).
base = imread(’klcc_a.png’);
unregistered = imread(’klcc_b.png’);

15. Specify control points in both images using cpselect (Figure 7.10). This is an
interactive process that is explained in detail in the IPT online documentation.
For the purpose of this tutorial, we will perform the following:
• Open the ​Control Point Selection ​tool.
• Choose a zoom value that is appropriate and lock the ratio.
• Select the ​Control Point Selection ​tool in the toolbar.
• Select a total of 10 control points per image, making sure that after we select
a point in one image with click on the corresponding point in the other image,
thereby establishing a match for that point. See Figure 7.11 for the points I
chose.
• Save the resulting control points using the ​File > Export Points to
Workspace ​option in the menu.
cpselect(unregistered, base);

16. Inspect the coordinates of the selected control points.

base_points
input_points
17. Use cpcorr to fine-tune the selected control points.
input_points_adj = cpcorr(input_points,base_points,...
unregistered(:,:,1),base(:,:,1))

18. This is a critical step.We need to specify the type of transformation we want to apply to the
unregistered image based on the type of distortion that it contains. In this case, since the
distortion appears to be a combination of translation, rotation, and scaling, we shall use the
’nonreflective similarity’ transformation type. This type requires only two pairs of control points.

19. Once we have selected the type of transformation, we can determine its parameters
using cp2tform.

20. Use the resulting tform structure to align the unregistered image (using
imtransform).
% Select the type of transformation
mytform1 = cp2tform(input_points,base_points,...
’nonreflective similarity’);
% Transform
% Transform the unregistered image
info = imfinfo(’klcc_a.png’);
registered = imtransform(unregistered,mytform1,...
’XData’,[1 info.Width], ’YData’,[1 info.Height]);
21. Display the registered image overlaid on top of the base image.
figure, imshow(registered);
hold on
h = imshow(base);
set(h, ’AlphaData’, 0.6)

4. 8.5 Tutorial 8.1: Gray-level Transformations / 163

1. Create an identity transformation function.
x = uint8(0:255);
plot(x); xlim([0 255]); ylim([0 255]);

2. Use the transformation function on the moon image to see how the identity
function works.
I = imread(’moon.tif’);
I_adj = x(I + 1);
figure, subplot(1,2,1), imshow(I), title(’Original Image’);
subplot(1,2,2), imshow(I_adj), title(’Adjusted Image’);
3. Create a negative transformation function and show the result after applied to
the moon image.
y = uint8(255:-1:0); I_neg = y(I + 1);
figure, subplot(1,3,1), plot(y), ...
title(’Transformation Function’), xlim([0 255]), ylim([0 255]);
subplot(1,3,2), imshow(I), title(’Original Image’);
subplot(1,3,3), imshow(I_neg), title(’Negative Image’);

4. Complement the original image and show that it is equivalent to the negative
image generated in the previous step.
I_cmp = imcomplement(I);
I_dif = imabsdiff(I_cmp,I_neg);
figure, imshow(I_cmp)
figure, imshow(I_dif,[])
5. Close all open figures and clear all workspace variables.
6. Generate a logarithmic transformation function.
x = 0:255; c = 255 / log(256);
y = c * log(x + 1);
figure, subplot(2,2,1), plot(y), ...
title(’Log Mapping Function’), axis tight, axis square
7. Use the transformation function to generate the adjusted image.
I = imread(’radio.tif’);
I_log = uint8(y(I + 1));
subplot(2,2,2), imshow(I), title(’Original Image’);
subplot(2,2,3), imshow(I_log), title(’Adjusted Image’);

8. Show a brightened version of the image.

I_br = imadd(I,100);
subplot(2,2,4), imshow(I_br), title(’Original Image Scaled’);

9. Use inverse log transformation to undo our previous transformation.

%
z = exp(x/c) - 1;
I_invlog = uint8(z(I_log + 1));
figure, subplot(2,1,1), plot(z), title(’Inverse-log Mapping Function’);
subplot(2,1,2), imshow(I_invlog), title(’Adjusted Image’);
10. Close all open figures and clear all workspace variables.
11. Generate an ​n​th root function where n equals 2.
x = 0:255; n = 2; c = 255 / (255 ˆ n);
root = nthroot((x/c), n);
figure, subplot(2,2,1), plot(root), ...
title(’2nd-root transformation’), axis tight, axis square

12. Use the transformation function to generate the adjusted image.

I = imread(’drill.tif’);
I_root = uint8(root(I + 1));
subplot(2,2,2), imshow(I), title(’Original Image’);
subplot(2,2,[3 4]), imshow(I_root), title(’Nth Root Image’);

13. Generate an ​n​th power transformation function.

power = c * (x .ˆ n);
figure, subplot(1,2,1), plot(power), ...
title(’2nd-power transformation’);
axis tight, axis square

14. Use the ​nt​ h power transformation to undo our previous transformation.
I_power = uint8(power(I_root + 1));
subplot(1,2,2), imshow(I_power), title(’Adjusted Image’);

15. Close all open figures and clear all workspace variables.
16. Load the micro image and display it.
I = imread(’micro.tif’);
figure, subplot(1,3,1), imshow(I), title(’Original Image’);

17. Create the transformation function.

y(1:175) = 0:174;
y(176:200) = 255;
y(201:256) = 200:255;
subplot(1,3,2), plot(y), axis tight, axis square
18. Generate the adjusted image.
I2 = uint8(y(I + 1));
subplot(1,3,3), imshow(I2), title(’Adjusted Image’);

19. Create a new transformation function and display the adjusted image.
z(1:175) = 50;
z(176:200) = 250;
z(201:256) = 50;
I3 = uint8(z(I + 1));
figure, subplot(1,2,1), plot(z), ...
xlim([0 255]), ylim([0 255]), axis square
subplot(1,2,2), imshow(I3)
20. Review the help information for glsdemo.
21. Run glsdemo with the image micro.tif and recreate the transformation
functions that we previously used in steps 17 and 19.

5.​​ 9.7 Tutorial 9.1: Image Histograms / 188

1. Display an image and its histogram.
I = imread(’circuit.tif’);
figure, subplot(2,2,1), imshow(I), title(’Image’)
subplot(2,2,2), imhist(I,256), axis tight, title(’Histogram’)
2. The previous step displayed the default histogram for the image—a histogram with 256 bins.
Let us see what happens if we change this value to 64 and 32.
subplot(2,2,3), imhist(I,64), axis tight, ...
title(’Histogram with 64 bins’)
subplot(2,2,4), imhist(I,32), axis tight, ...
title(’Histogram with 32 bins’)

3. Get the values of each bin in the histogram for later use.
c = imhist(I,32);

4. Normalize the values in c.

c_norm = c / numel(I);
5. Close any open figures.
6. Display the histogram data using a bar chart.
figure, subplot(1,2,1), bar_1 = bar(c);
set(gca, ’XLim’, [0 32], ’YLim’, [0 max(c)]);

7. Set the tick marks to reflect the limits of the graph.

set(gca, ’XTick’, [0:8:32], ’YTick’, ...
[linspace(0,7000,8) max(c)]);
8. Use the set function to change the color of the bar chart. Also, give the chart
a title.
set(bar_1, ’FaceColor’, ’r’), title(’Bar Chart’)

9. Display the normalized bar chart and customize its display.

subplot(1,2,2), bar_2 = bar(c_norm);
set(gca, ’XTick’, [0:8:32], ’YTick’, ...
[linspace(0,0.09,10) max(c_norm)])
xlim([0 32]), ylim([0 max(c_norm)])
title(’Normalized Bar Chart’)
set(bar_2, ’FaceColor’, ’g’)

10. Close any open figures.

11. Display stem charts for both standard and normalized histogram data.
figure,
subplot(1,2,1), stem(c,’fill’,’MarkerFaceColor’,’red’), ...
axis tight, title(’Stem Chart’)
subplot(1,2,2), stem(c_norm,’fill’,’MarkerFaceColor’,’red’), ...
axis tight, title(’Normalized Stem Chart’)

12. Display a plot graph for both standard and normalized histogram data.
figure, subplot(1,2,1), plot(c), axis auto, title(’Plot Graph’)
subplot(1,2,2), plot(c_norm), axis auto, ...
title(’Normalized Plot Graph’)

9.8 Tutorial 9.2: Histogram Equalization and Specification / 191

1. Display the image pout and its histogram.
I = imread(’pout.tif’);
figure, subplot(2,2,1), imshow(I), ...
title(’Original Image’)
subplot(2,2,2), imhist(I), ...
title(’Original Histogram’)

2. Use the histeq function to perform histogram equalization.

I_eq = histeq(I,256);

3. Display the equalized image and its histogram.

subplot(2,2,3), imshow(I_eq), title(’Equalized Image’)
subplot(2,2,4), imhist(I_eq), title(’Equalized Histogram’)
4. Close any open figures and clear all workspace variables.
5. Execute the following code to perform histogram equalization on the tire
image.
I = imread(’tire.tif’); I_eq = histeq(I,256);
figure, subplot(2,2,1), imshow(I), title(’Original Image’)
subplot(2,2,2), imhist(I), title(’Original Histogram’)
subplot(2,2,3), imshow(I_eq), title(’Equalized Image’)
subplot(2,2,4), imhist(I_eq), title(’Equalized Histogram’)

6. Close any open figures and clear all workspace variables.

7. Perform histogram equalization on the eight image.
I = imread(’eight.tif’); I_eq = histeq(I,256);
figure, subplot(2,2,1), imshow(I), title(’Original Image’)
subplot(2,2,2), imhist(I), title(’Original Histogram’)
subplot(2,2,3), imshow(I_eq), title(’Equalized Image’)
subplot(2,2,4), imhist(I_eq), title(’Equalized Histogram’)
8. Display the normalized cdf for the eight.tif image.
I_hist = imhist(I); tf = cumsum(I_hist); tf_norm = tf / max(tf);
figure, plot(tf_norm), axis tight

9. The transformation function can also be obtained without using the cumsum
function.
[newmap, T] = histeq(I);
figure, plot(T)

10. Close any open figures and clear all workspace variables.
11. Prepare a subplot and display original image and its histogram.
img1 = imread(’pout.tif’);
figure, subplot(3,3,1), imshow(img1), title(’Original Image’)
subplot(3,3,2), imhist(img1), title(’Original Histogram’)
12. Display the image after histogram equalization for comparison.
img1_eq = histeq(img1); m1 = ones(1,256)*0.5;
subplot(3,3,4), imshow(img1_eq), title(’Equalized Image’)
subplot(3,3,5), imhist(img1_eq), title(’Equalized Histogram’)
subplot(3,3,6), plot(m1), title(’Desired Histogram Shape’), ...
ylim([0 1]), xlim([1 256])

13. Display matched image where the desired histogram shape is a straight line
from (0, 0) to (1, 1).
m2 = linspace(0,1,256); img2 = histeq(img1,m2);
subplot(3,3,7), imshow(img2), title(’Matched Image’)
subplot(3,3,8), imhist(img2), title(’Matched Histogram’)
subplot(3,3,9), plot(m2), title(’Desired Histogram Shape’), ...
ylim([0 1]), xlim([1 256])
14. Close any open figures and clear all workspace variables.
15. Run the ​Interactive Histogram Matching d​ emo.
Ihmdemo
16. Experiment with creating your own desired histogram shape. To create new
points on the function curve, click the curve at the desired location. To move
a point, press and drag the point. To delete a point, simply click it.
17. Perform local histogram equalization on the coins image.

9.9 Tutorial 9.3: Other Histogram Modification Techniques / 195

1. Display original image and prepare subplot.
2. Obtain a brighter version of the input image by adding 0.1 to each pixel

3. Produce another brighter image by adding 0.5 to original image.

4. Close any open figures.

5. Execute the following code to see histogram stretching on the pout image,
which is already loaded in variable I.
6. Perform histogram stretching with imadjust using default parameters and
confirm that the results are identical to the ones obtained before

7. Close any open figures and clear all workspace variables.

8. Execute the following code to see the result of histogram shrinking.

9. Display the transformation function for the adjustment performed in the previous
step.
10. Close any open figures.
11. Perform histogram shrinking with a gamma value of 2.
6.10.7 Tutorial 10.1: Convolution and Correlation / 223
Specify the two matrices to be used.
a = [0 0 0 1 0 0 0];
f = [1 2 3 4 5];

2. Perform convolution, using a as the input matrix and f as the filter.

g = imfilter(a,f,’full’,’conv’)
3. Perform correlation on the same set of matrices.
h = imfilter(a,f,’full’,’corr’)

4. Clear all workspace variables.

5. Use imfilter to perform a correlation of the two matrices
x = [140 108 94;89 99 125;121 134 221]
y = [-1 0 1;-2 0 2;-1 0 1]
z = imfilter(x,y,’corr’)
6. Use imfilter to perform a convolution of the two matrices.
z2 = imfilter(x,y,’conv’)

10.8 Tutorial 10.2: Smoothing Filters in the Spatial Domain / 225

1. Load the cameraman image and prepare a subplot.
2. Create a mean (averaging) filter automatically through thefspecial function

3. Filter the cameraman image with the generated mask.

4. Create a nonuniform version of the mean filter.

5. Filter the original image with the new, nonuniform averaging mask.
6. Create a Gaussian filter and display the kernel as a 3D plot.

7. Filter the cameraman image using the Gaussian mask.

10. Tutorial 10.3: Sharpening Filters in the Spatial
9 Domain / 228

1. Load the moon image and prepare a subplot figure

2. Create a Laplacian kernel and apply it to the image using the imfilter
function.

3. Display a scaled version of the Laplacian image for display purposes

4. Subtract the filtered image from the original image to create the sharpened image.

5. Use the composite Laplacian mask to perform image sharpening in one step
6. Close all open figures and clear all workspace variables.
7. Load the moon image and generate the blurred image

8. Shrink the histogram of the blurred image

9. Now subtract the blurred image from the original image

10. Stretch the sharpened image histogram to the full dynamic grayscale range and display the
final result.

11. Subtract the blurred image from the original image to generate a sharpening image.
12. Add sharpening image to original image to produce the final result.

13. Generate unsharp masking kernel using the fspecial function

14. Apply the mask to the original image to create a sharper image.
15. Close any open figures.
16. Create a high-boost mask (where A = 1) and apply it to the moon image

17. Show that a high-boost mask when A = 3 looks similar to the image simply multiplied by 3.
7.11.5 Tutorial 11.1: 2D Fourier Transform / 252
1. Load the cameraman image, convert it to double (one of the data classes
accepted as an input to fft2), and generate its FT.

2. Shift the FT array of results.

3. Display the FT results, remapped to a grayscale range.

4. Display the log of the shifted FT image

5. Close any open figures.

6. Generate a distance matrix that is the same size as the image I.
[M, N] = size(I);
D = distmatrix(M, N);
7. Create a 3D mesh plot of the distance matrix.
D_shift = fftshift(D);
figure, mesh(D_shift
8.12.6 Tutorial 12.1: Noise Reduction Using Spatial-domain
Techniques / 289
1. Load the eight image, add (Gaussian) noise to it, and display the image before
and after adding noise.

2. Apply an averaging filter to the image using the default kernel size (3 × 3).
f1 = fspecial(​'average'​);
I_blur1 = imfilter(In,f1);

3. Implement an averaging kernel with a 5 × 5 mask.

5. Load two noisy versions of the eight image: one with salt noise and the other with
pepper. Also, display the original image along with the two affected images

6. Filter the salt noise affected image using −1 for the value of r.
I_fix1 = nlfilter(I_salt,[3 3],@c_harmonic,-1);
subplot(2,3,5), imshow(I_fix1), title(’Salt Removed, r = -1’);

9.13.8 Tutorial 13.1: Binary Morphological Image Processing / 325

1. Load and display the blobs test image
2. Create a 3×3 structuring element with all coefficients equal to 1.

3. Perform dilation using the generated SE and display the results.

4. Create a 1 × 7 SE with all elements equal to 1 and dilate the image.

5. Erode the original image with a 3 × 3 structuring element and display the results.
6.Erode the original image with a 1 × 7 structuring element

7. Perform morphological opening on the original image using the square 3 × 3

SE created previously
8.Compare opening with eroding.

9. Open the image with a 1 × 7 SE.

10. Create a square 5 × 5 SE and perform morphological closing on the image

11.

12. Close any open windows.

13. Define the two structuring elements.
SE1 = [ 0 0 0 0 0
00000
01100
00100
0 0 0 0 0]
SE2 = [ 0 0 0 0 0
11110
00010
00010
0 0 0 1 0]
14. Apply the HoM operation on the original image
13.9 Tutorial 13.2: Basic Morphological Algorithms / 330
1. I = imread(’morph.bmp’);
figure, imshow(I), title(’Original image’);
2. Subtract the original image from its eroded version to get the boundary image

3. Perform boundary extraction using the ​bwperim ​function.

5. Fill holes in the image using the ​imfill ​function.

I_fill1 = imfill(I,’holes’);
figure, imshow(I_fill1), title(’Holes filled’);
6. Pick two of the three holes interactively by executing this statement. After
selecting the points, press Enter.

9. Label the objects in an image using ​bwlabel​.

10. Use ​bwmorph ​to thin the original image with five iterations.
11. Thicken the original image with five iterations.

10. 14.7 Tutorial 14.1: Edge Detection / 354 

1.​ Load and display the test image. 
  
2. Extract the edges in the image using the Prewitt operator

 
 
 
 
4. Extract the edges from the test image using the Sobel edge detector. 
  
5. Extract the edges from the test image with Gaussian noise using the Sobel edge
detector.
6. Extract the edges from the test image with the Sobel operator with no thinning

 
 
 
 11. 15.5 Tutorial 15.1: Image Thresholding / 379 
1. Load and display the test image 

2. Display a histogram plot of the ​coins ​image to determine what threshold level
to use.

4. Set the threshold value to 85 and generate the new image.

5. Use the ​graythresh ​function to generate the threshold value automatically

7.Load the ​gradient_with_text ​image and prepare a subplot

8. Globally threshold the image

11. Define the function as well as its input and output parameters in the first line.
function y = adapt_thresh(x)
12. Add this line of code under the function definition.
y = im2bw(x,graythresh(x));

14. Perform adaptive thresholding by entering the following command in the command
window. Note that it may take a moment to perform the calculation, so
be patient.
I_thresh = blkproc(I,[10 10],@adapt_thresh);
15. Display the original and new image.
 12. 16.6 Tutorial 16.1: Pseudocolor Image Processing / 419
1. Load the ​onions.png ​image and display its RGB components

2. Convert the image to HSV and display its components.

3. Convert the original image to grayscale and compare it with the value component
of the HSV image.

4. Apply a smoothing filter to each component and then reconstruct the image