CS4509- Digital image processing Laboratory Department of ADS 2023-2024

Vision of the Department:

To emerge as a Centre of Excellence in the field of Artificial Intelligence and to accomplish
eminence to have global recognition through education, innovation and collaborative
research in the realm of Data Science.

Mission of the Department:

 To impart quality value-based technical education in cutting edge Artificial
Intelligence technologies that bridge the gap between academia and industry.
 To promote research, innovation and entrepreneurial skills with the latest
technologies to be globally competitive professionals.
 To equip aspiring students with the skills to perform intelligent data analysiswhich in
turn provide dynamic and lucrative careers in the field of technology.
 To develop technocrats with strong core capabilities in Artificial Intelligence by
providing good infrastructure, resources, effective teaching learning process and
competent with state-of-the-art technologies.

ANNEXURE I
PROGRAM OUTCOMES (POs)
Engineering graduates will be able to:
1. Engineering Knowledge: Apply the knowledge of mathematics, science, engineering
fundamentals, and an engineering specialization to the solution of complex engineering
problems.
2. Problem analysis: Identify, formulate, review research literature, and analyze complex
engineering problems reaching substantiated conclusions using first principles of
mathematics, natural sciences, and engineering sciences.
3. Design/development of solutions: Design solution for complex engineering problems
and design systems components or process that meet the specified needs with
appropriate consideration for the public health and safety, and the cultural, societal, and
environmental considerations.
4. Conduct investigations of complex problems: Use research- based knowledge and
research methods including design of experiments, analysis and interpretation of data,
and synthesis of the information to provide valid conclusions.
5. Modern tool usage: Create, select, and apply appropriate techniques, resources, and
modern engineering and IT tools including prediction and modeling to complex
engineering activities with an understanding of the limitations.
6. The engineer and society: Apply reasoning informed by the contextual knowledge to
assess societal, health, safety, legal and cultural issues and the consequent
responsibilities relevant to the professional engineering practice.
7. Environmental and sustainability: Understand the impact of the professional
engineering solutions in societal and environmental contexts and demonstrate the
knowledge of, and need for sustainable development.
8. Ethics: Apply ethical principles and commit to professional ethics and responsibilities
and norms of the engineering practice.
9. Individual and team work: Function effectively as an individual, and as a member or
leader in diverse teams, and in multidisciplinary settings.
10. Communication: Communicate effectively on complex engineering activities with the
St. Joseph’s Institute of Technology 1
CS4509- Digital image processing Laboratory Department of ADS 2023-2024

engineering community and with society at large, such as, being able to comprehend
and write effective reports and design documentation, make effective presentations, and
give and receive clear instructions.
11. Project management and finance: Demonstrate knowledge and understanding of the
engineering and management principles and apply these to one’s own work, as a
member and leader in a team, to manage projects and in multidisciplinary
environments.
12. Life-Long learning: Recognize the need for, and have the preparation and ability to
engage in independent and life-long learning in the broadest context of technological
change.

PROGRAM EDUCATIONAL OBJECTIVES (PEOs)

PEO1: To Build next generation of highly skilled graduates with a strong knowledge in
ArtificialIntelligence and Data Science to contribute and innovate new technologies for societal
needs.
PEO2: To Create Engineers to promote collaborative learning and to exhibit their
employability skills and practice the ethics of their profession through innovation or
entrepreneurship.
PEO3: To Pursue graduate studies in the field of Data Science and to be committed in lifelong
researchtowards social, political, and technical issues.
PEO4: To Exhibit innovative thoughts in Engineering, Problem Solving and Critical Thinking
skills toexcel in interdisciplinary domains.

PROGRAM SPECIFIC OUTCOMES (PSOs)

PSO 1: To understand, analyze and apply the AI based efficient domain specific processes for
problem- solving, inference, perception, knowledge representation and learning to design
computer-based systems for varying complexity.
PSO 2: To implement search algorithms, neural networks, machine learning and data analytics
to create innovative solutions from idea to product for successful career and entrepreneurship.
PSO 3: To develop intelligent solutions and project development skills using Data Science
technologies to cater to the societal needs.
PSO 4: To provide a concrete foundation and enrich their abilities to qualify for Employment,
Higher studies and Research in Artificial Intelligence and Data Science with ethical values.

Course Outcomes (COs):

CO1 To understand image acquisition and analyze different image transforms on images

CO2 To understand image enhancement techniques

CO3 To understand segmentation and compression techniques

St. Joseph’s Institute of Technology 2

CS4509- Digital image processing Laboratory Department of ADS 2023-2024

Attainment of POs through COs (Legends: 1 – Low, 2 – Medium, 3 – High)

CO PO PSO
PO PO PO1 PSO PSO PSO
PO1 PO2 PO3 PO4 PO5 PO6 PO7 PO8 PO9
10 11 2 1 2 3
CO1 3 3 3 1 1 - - 2 2 2 - 2 3 3 2
CO2 3 3 3 1 1 - - 2 2 2 - 2 3 3 2
CO3 2 3 3 1 1 - - 2 2 2 - 2 3 3 2
Relation between Course content / Experiments with CO

S. Knowledge Course
Experiment
No level Outcomes
1. BL3 To acquire and Display of an Image, Negative of an Image CO1
(Binary & Gray Scale)
2. BL3 Implementation of Relationships between Pixels CO1
3. BL3 Analysis of images with different color models CO1
4. BL3 Implementation of Transformations of an Image CO1
5. BL3 Histogram Processing and Basic Thresholding functions CO2
6. BL3 Computation of Mean, Standard Deviation, Correlation CO2
coefficient of the given Image
7. BL3 Implementation of Image Enhancement-Spatial filtering CO2
8. BL3 Implementation of Image Enhancement- Filtering in frequency CO2
domain
9. BL4 Image segmentation – Edge detection, line detection and point CO3
detection.
10. BL4 Implementation of Region based Segmentation CO3
11. BL4 Basic Morphological operations. CO3
12. BL4 Implementation of Image compression techniques CO3
BL1-Remember, BL2-Apply, BL3-Analyze, BL4-Understand, BL5-Evaluate, BL6-Create

St. Joseph’s Institute of Technology 3

CS4509- Digital image processing Laboratory Department of ADS 2023-2024

INDEX

Name of the Student: Register Number:

Roll Number: Department:
Name of the Lab-Incharge:

Date of Page
S.No Date Name of the Experiment Staff Sign Remarks
Submission No.

St. Joseph’s Institute of Technology 4

CS4509- Digital image processing Laboratory Department of ADS 2023-2024

ANNA UNIVERSITY SYLLABUS

CS4509 DIGITAL IMAGE PROCESSING LABORATORY LTPC0042

LIST OF EXPERIMENTS
1. To acquire and Display of an Image, Negative of an Image (Binary & Gray Scale)
2. Implementation of Relationships between Pixels
3. Analysis of images with different color models.
4. Implementation of Transformations of an Image
5. Histogram Processing and Basic Thresholding functions
6. Computation of Mean, Standard Deviation, Correlation coefficient of the given Image
7. Implementation of Image Enhancement-Spatial filtering
8. Implementation of Image Enhancement- Filtering in frequency domain
9. Image segmentation – Edge detection, line detection and point detection.
10. Implementation of Region based Segmentation
11. Basic Morphological operations.
12. Implementation of Image compression techniques

CYCLE – I
1. To acquire and Display of an Image, Negative of an Image (Binary & Gray Scale)
2. Implementation of Relationships between Pixels
3. Analysis of images with different color models.
4. Implementation of Transformations of an Image
5. Histogram Processing and Basic Thresholding functions
6. Computation of Mean, Standard Deviation, Correlation coefficient of the given Image

CYCLE - II

1. Implementation of Image Enhancement-Spatial filtering

2. Implementation of Image Enhancement- Filtering in frequency domain
3. Image segmentation – Edge detection, line detection and point detection.
4. Implementation of Region based Segmentation
5. Basic Morphological operations.
6. Implementation of Image compression techniques

St. Joseph’s Institute of Technology 5

CS4509- Digital image processing Laboratory Department of ADS 2023-2024

Ex. No. ACQUIRE AND DISPLAY AN IMAGE, NEGATIVE OF Date

AN IMAGE (BINARY & GRAY SCALE)
AIM:
The aim is to acquire an image, convert it into its negative, and display both the original and
negative images in both binary and grayscale formats using MATLAB.

SOFTWARE REQUIRED:
MATLAB 2013b

THEORY:
Image Acquisition: To acquire an image, you can use the `imread` function in MATLAB.
This function reads an image file and stores it as a matrix. The image can be in various formats
such as JPEG, PNG, or BMP.
Negative of an Image:
For Grayscale Images: The negative of a grayscale image can be obtained by subtracting each
pixel value from the maximum intensity value. For an 8-bit image, this maximum value is 255.
For Binary Images: In binary images, where pixel values are typically 0 or 1, the negative is
obtained by swapping 0s with 1s and vice versa.
Displaying Images: You can use the `imshow` function in MATLAB to display images. For
binary images, you may want to adjust the colormap to have a clear representation of 0s and
1s.

PROCEDURE:
1. Image Acquisition - Read the original image with the image filename
2. Negative of an Image:
a) Convert the original image to grayscale if needed
b) Calculate the negative of the grayscale image
c) For binary images, you should threshold the image first
d) Assuming your binary image is already threshold
e) If not, use an appropriate thresholding technique
f) Adjust the threshold value as needed
g) Calculate the negative of the binary image
3.Displaying Images:
a) Display the original grayscale image
b) Display the negative grayscale image
c) Display the original binary image
d) Display the negative binary image
4. Save the displayed images to files using `imwrite` if needed.
PROGRAM:
% Red Blue and Green and Gray Components
i=imread('cancercell.jpg');
subplot(3,2,1); imshow(i); title('Original Image');
%Red Component
r=i(:,:,1);
subplot(3,2,2); imshow(r);title('Red Component');
%Green Component
g=i(:,:,2);
subplot(3,2,3); imshow(g); title('Green Component');
St. Joseph’s Institute of Technology 6
CS4509- Digital image processing Laboratory Department of ADS 2023-2024

%Blue Component
b=i(:,:,3);
subplot(3,2,4); imshow(b); title('Blue Component');
%Color to Gray Image
rg=rgb2gray(i);
subplot(3,2,5); imshow(rg); title('Gray Image');
Complement, Converting and Simulation of an Image
% Display color Image, find its complement and convert to gray scale
I=imread('cancercell.jpg');
subplot(2,2,1); imshow(I); subimage(I);
title('Color Image');
c=imcomplement(I);
subplot(2,2,2); imshow(c); subimage(c);
title('Complement of color Image');
r=rgb2gray(I);
subplot(2,2,3); imshow(r); subimage(r);
title('Gray scale of color Image');
%Complement of Gray Image
b=imcomplement(r);
subplot(2,2,4); imshow(b); subimage(b);
title('Complement of Gray Image');
%Simulation of an Image (Arithmetic & Logic Operation)
a=ones(40); b=zeros(40); c=[a b;b
a]; d=[b b;a a]; A=10*(c+d);
M=c.*d;
S=c-d; D=c/4; figure;
subplot(3,2,1); imshow(c);
subplot(3,2,2); imshow(d);
subplot(3,2,3); imshow(A);
subplot(3,2,4); imshow(M);
subplot(3,2,5); imshow(S);
subplot(3,2,6); imshow(D);

OUTPUT:

St. Joseph’s Institute of Technology 7

CS4509- Digital image processing Laboratory Department of ADS 2023-2024

RESULT:

REVIEW QUESTIONS:
1. What MATLAB function is used to acquire an image from a file?
2. Why might it be necessary to convert a colour image to grayscale before processing it
further?
3. How is the negative of a grayscale image calculated, and what is the significance of the
value 255 in this context?
4. In binary image processing, how is the negative of a binary image typically obtained?
5. Can you describe a scenario where calculating the negative of an image is useful in image
processing?

Ex. No. IMPLEMENTATION OF RELATIONSHIPS Date

BETWEEN PIXELS
AIM:
The aim is to implement relationships between pixels in a digital image.

SOFTWARE REQUIRED:
MATLAB 2013b

THEORY:
Relationships between pixels in an image refer to the interactions and dependencies among
neighboring or non-neighboring pixels. These relationships are crucial for tasks such as noise
reduction, feature extraction, object detection, and more. MATLAB represents an image as a
matrix, with each element corresponding to a pixel. The central pixel's location is typically
St. Joseph’s Institute of Technology 8
CS4509- Digital image processing Laboratory Department of ADS 2023-2024

denoted by its row and column coordinates in the matrix. MATLAB represents an image as a
matrix, with each element corresponding to a pixel. To find a pixel's neighbors, we can specify
a neighborhood size, such as a square or circular region, centered on the pixel of interest.
MATLAB's array indexing allows for easy extraction of the neighboring pixels.

PROCEDURE:
1. Load the image into MATLAB.
2. Use the `imread` function in MATLAB to read the image.
3. Determine the specific relationship or operation you want to implement.
4. Implement the chosen relationship between pixels. Calculate the new pixel values based
on a mapping function that stretches the intensity values.
5. Create a kernel or mask for the specific filtering operation (e.g., Gaussian, Sobel, or
custom filters). Use convolution to apply the filter to the image.
6. For segmentation tasks, choose an appropriate method such as thresholding, region-
growing, or edge-based segmentation.
7. Implement the chosen method to divide the image into distinct regions or objects.
8. Display the original image alongside the processed image to visualize the impact of the
implemented relationship or operation.
9. Testing and Evaluation: Test the implemented relationship or operation on different
images and evaluate its effectiveness for the intended image processing task.

PROGRAM:
% To find Neighbour of a given Pixel
a=magic(5);
disp(‘a=’); disp(a);
b=input('Enter the row < size of the Matrix');
c=input(' Enter the Column < size of matrix');
disp(‘Element’); disp(a(b,c));
% 4 Point Neighbour
N4=[a(b+1,c), a(b-1,c), a(b,c+1), a(b,c-1)];
disp(‘N4=’); disp(N4);
%8 Point Neighbour
N8=[a(b+1,c), a(b-1,c), a(b,c+1), a(b,c-1), a(b+1,c+1), a(b+1,c-1), a(b-1,c-1), a(b-1,c+1)];
disp(‘N8=’); disp(N8);
%Diagonal Neighbour
ND=[ a(b+1,c+1), a(b+1,c-1), a(b-1,c-1), a(b-1,c+1)];
disp(‘ND=’); disp(ND);

OUTPUT:

>> pixel
17 24 1 8 15
23 5 7 14 16
4 6 13 20 22
10 12 19 21 3
11 18 25 2 9

Enter the row < size of the Matrix 3

St. Joseph’s Institute of Technology 9

CS4509- Digital image processing Laboratory Department of ADS 2023-2024

Enter the Column < size of matrix 2

N4:
12 5 13 4

N8:
12 5 13 4 19 10 23 7

ND:
19 10 23 7

RESULT:

REVIEW QUESTIONS:
1. What does it mean to implement relationships between pixels in image processing?
2. Give an example of an image processing task that relies on defining and using
relationships between pixels.
3. How would you implement a pixel-wise contrast stretching operation in MATLAB, and
why might this be useful in enhancing images?
4. What is image filtering, and why is it an essential technique in image processing?
5. Describe the process of applying a convolution filter to an image and provide an
example of a filter that can be used for a specific purpose.

Ex. No. ANALYSIS OF IMAGES WITH DIFFERENT Date

COLOR MODELS
AIM: The aim is to analyze images using different color models, such as RGB, HSL, and
CMYK, to gain insights into color-based features and properties within the images.

SOFTWARE REQUIRED:
MATLAB 2013b

THEORY:
Color models are mathematical representations of colors in digital images. Common color
models include RGB (Red, Green, Blue), HSL (Hue, Saturation, Lightness), and CMYK (Cyan,
Magenta, Yellow, Key/Black).
In the RGB model, colors are represented by combining various intensities of red, green, and
blue. It is the most common color model used in digital displays and cameras.
The HSL model focuses on human perception of colors. It represents colors based on their hue,
saturation, and lightness. This model is valuable for color manipulation and image analysis.
CMYK is used primarily in color printing and subtractive color mixing. It represents colors
using cyan, magenta, yellow, and key (black) components. Analyzing images in CMYK can
help with color correction and print-related tasks.

PROCEDURE:
1. Load the image in MATLAB.
2. Convert the loaded image to different color models, such as RGB, HSL, and CMYK, using
built-in functions or libraries. In MATLAB, you can use functions like `rgb2hsl` or
`rgb2cmyk`.
St. Joseph’s Institute of Technology 10
CS4509- Digital image processing Laboratory Department of ADS 2023-2024

3. Calculate statistical metrics for each color channel in the chosen color model. These metrics
may include mean, standard deviation, or histograms of pixel values.
4. Display the image in each color model to visually compare and understand how the image's
appearance changes.
5. Apply color manipulations or corrections within the chosen color model to enhance or
modify specific aspects of the image, such as brightness, contrast, or color balance.
6. Extract color-based features from the different color channels, which can be used for further
analysis or object recognition.
7. Perform image segmentation using color information from the selected color model to
separate objects or regions of interest in the image based on color.
8. Display the results of your analysis, such as histograms of color channels, segmented
regions, or color-corrected images.

PROGRAM:
% Load an image
originalImage = imread('image.jpg'); % Replace 'image.jpg' with the image filename

% Display the original image

figure;
subplot(2, 2, 1);
imshow(originalImage);
title('Original Image');

% Convert the image to RGB

rgbImage = originalImage;

% Display the RGB image

subplot(2, 2, 2);
imshow(rgbImage);
title('RGB Image');

% Conversion to HSL (Custom function required)

% hslImage = rgb2hsl_custom(rgbImage);

% Display the HSL image

% subplot(2, 2, 3);
% imshow(hsl2rgb_custom(hslImage));
% title('HSL Image');

% Conversion to CMYK (Custom function required)

% cmykImage = rgb2cmyk_custom(rgbImage);

% Display the CMYK image

% subplot(2, 2, 4);
% imshow(cmyk2rgb_custom(cmykImage));
% title('CMYK Image');

% Analyze color-based features

St. Joseph’s Institute of Technology 11

CS4509- Digital image processing Laboratory Department of ADS 2023-2024

% Dominant colors in RGB

rgbDominantColor = calculateDominantColor(rgbImage);
fprintf('Dominant color in RGB: R=%d, G=%d, B=%d\n', rgbDominantColor(1),
rgbDominantColor(2), rgbDominantColor(3));

% % Dominant colors in HSL (Custom function required)

% hslDominantColor = calculateDominantColor(hsl2rgb_custom(hslImage));
% fprintf('Dominant color in HSL: R=%d, G=%d, B=%d\n', hslDominantColor(1),
hslDominantColor(2), hslDominantColor(3));

% % Dominant colors in CMYK (Custom function required)

% cmykDominantColor = calculateDominantColor(cmyk2rgb_custom(cmykImage));
% fprintf('Dominant color in CMYK: R=%d, G=%d, B=%d\n', cmykDominantColor(1),
cmykDominantColor(2), cmykDominantColor(3));

% Calculate and plot color histograms

rgbHistogram = calculateColorHistogram(rgbImage);
% hslHistogram = calculateColorHistogram(hsl2rgb_custom(hslImage));
% cmykHistogram = calculateColorHistogram(cmyk2rgb_custom(cmykImage));

% Plot RGB histogram

figure;
subplot(3, 1, 1);
bar(rgbHistogram, 'r');
title('RGB Histogram');

% % Plot HSL histogram

% subplot(3, 1, 2);
% bar(hslHistogram, 'g');
% title('HSL Histogram');

% % Plot CMYK histogram

% subplot(3, 1, 3);
% bar(cmykHistogram, 'b');
% title('CMYK Histogram');

% Function to calculate dominant color (implementation required)

% % Custom RGB to HSL conversion function

% function hslImage = rgb2hsl_custom(rgbImage)
% % Implement RGB to HSL conversion here
% end

% % Function to calculate dominant color

function dominantColor = calculateDominantColor(image)
% Convert the image to double precision
doubleImage = im2double(image);

% Reshape the image into a 2D array of RGB values

St. Joseph’s Institute of Technology 12
CS4509- Digital image processing Laboratory Department of ADS 2023-2024

[height, width, ~] = size(doubleImage);

reshapedImage = reshape(doubleImage, height * width, 3);

% Perform k-means clustering (change k to the desired number of clusters)

k = 5; % You can adjust this value
[clusterIndices, ~] = kmeans(reshapedImage, k);

% Find the largest cluster (i.e., the dominant color)

clusterCounts = hist(clusterIndices, 1:k);
[~, dominantCluster] = max(clusterCounts);

% Calculate the RGB values of the dominant color

dominantColor = mean(reshapedImage(clusterIndices == dominantCluster, :), 1);
end

% % Function to calculate color histograms

function colorHistogram = calculateColorHistogram(image)
% Convert the image to double precision
doubleImage = im2double(image);

% Separate the RGB channels

redChannel = doubleImage(:, :, 1);
greenChannel = doubleImage(:, :, 2);
blueChannel = doubleImage(:, :, 3);

% Define the number of bins for the histogram

numBins = 256; % You can adjust this value as needed

% Calculate histograms for each channel

redHist = imhist(redChannel, numBins);
greenHist = imhist(greenChannel, numBins);
blueHist = imhist(blueChannel, numBins);

% Combine the individual histograms into a single histogram

colorHistogram = [redHist, greenHist, blueHist];

% Normalize the histogram to have values between 0 and 1

colorHistogram = colorHistogram / sum(colorHistogram);

% Plot the color histogram (optional)

figure;
subplot(2, 1, 1);
bar(colorHistogram);
title('Color Histogram');

% Plot individual channel histograms (optional)

subplot(2, 1, 2);
bar(redHist, 'r');
hold on;
St. Joseph’s Institute of Technology 13
CS4509- Digital image processing Laboratory Department of ADS 2023-2024

bar(greenHist, 'g');
bar(blueHist, 'b');
title('Channel Histograms');
end

OUTPUT:

Dominant color in RGB: R=1.290451e-01, G=1.700964e-01, B=1.570514e-01

St. Joseph’s Institute of Technology 14

CS4509- Digital image processing Laboratory Department of ADS 2023-2024

RESULT:

REVIEW QUESTIONS:
1. What are the primary components of the RGB color model, and how are colors
represented in this model?
2. Explain the significance of the HSL color model, including its components: Hue,
Saturation, and Lightness.
3. How can you extract the hue channel from an image in the HSL color model using
MATLAB, and what kind of image analysis tasks might benefit from this?
4. Describe the purpose and usage of the CMYK color model, particularly in the context of
color printing.
5. What is the process in MATLAB for converting an image from RGB to CMYK, and how
does it contribute to color correction and print-related tasks?

Ex. No. IMPLEMENTATION OF TRANSFORMATIONS Date

OF AN IMAGE

AIM: The aim of experiment is to implement transformations on digital images, such as

resizing, rotation, and flipping, to alter the appearance and structure of the image while
maintaining image quality.

SOFTWARE REQUIRED:
MATLAB 2013b

THEORY:
Image transformations are operations that modify the spatial and visual properties of digital
images. Common transformations include resizing, rotation, cropping, and flipping. Image
resizing involves changing the dimensions of the image, either by scaling it up (enlarging) or
down (shrinking). This transformation is useful for adjusting image dimensions or preparing
images for different display or printing sizes. Image rotation is the process of changing the
image's orientation by a specific angle, often 90, 180, or 270 degrees. It is used to correct image
alignment or for artistic effects. Image flipping includes horizontal and vertical mirroring,
which reverses the image along the horizontal or vertical axis. Flipping is valuable for creating
mirror images or altering the image's perspective.

St. Joseph’s Institute of Technology 15

CS4509- Digital image processing Laboratory Department of ADS 2023-2024

PROCEDURE:
1. Load the digital image you want to transform using MATLAB.
2. Implement resizing by specifying the new dimensions or scaling factor. Interpolation
methods, such as nearest-neighbor or bilinear, can be used to fill in pixel values during
resizing.
3. Rotate the image by a specified angle, typically using transformation matrices. Be
mindful of the interpolation method to maintain image quality during rotation.
4. Perform horizontal or vertical flipping to mirror the image. This involves rearranging
pixel values along the specified axis.
5. Display the original image and the transformed image to visually compare the effects of
the applied transformations.

PROGRAM:

%Scaling & Rotation

% Scaling (Resize)
I=imread('earcell.jpg');
subplot(2,2,1); subimage(I); title('Original Image');
s=input('Enter Scaling Factor');
j=imresize(I,s);
subplot(2,2,2); subimage(j); title('Scaled Image');
% Rotation
K=imrotate(j,60);
subplot(2,2,3); imshow(K); title('Rotated Image 60deg');
R=imrotate(j,45);
subplot(2,2,4); imshow(R); title('Rotated Image 45deg');
%Display the color image and its Resized images by different methods
%Display the color image
I=imread('embryo.jpg'); figure,
subplot(2,2,1);
subimage(I);
title('Original Image');
%Display Resized image by Bilinear method
B=imresize(I,5);
subplot(2,2,2); subimage(B);
title('Bilinear Image');
%Display Resized image by Nearest method
C=imresize(I,5,'nearest');
subplot(2,2,3); subimage(C);
title('Nearest Image');
%Display Resized image by Bicubic method
D=imresize(I,5,'Bicubic');
subplot(2,2,4); subimage(D);
title('Bicubic Image');

St. Joseph’s Institute of Technology 16

CS4509- Digital image processing Laboratory Department of ADS 2023-2024

OUTPUT:

RESULT:

St. Joseph’s Institute of Technology 17

CS4509- Digital image processing Laboratory Department of ADS 2023-2024

REVIEW QUESTIONS:
1. What is the purpose of image resizing, and how does it affect the image's dimensions?
2. Explain the role of interpolation methods in image resizing, and provide an example of a
scenario where resizing is useful.
3. How can you ensure that image quality is maintained when rotating an image in MATLAB?
4. Give an example of an application where image flipping is beneficial.
5. What factors should be considered when assessing the quality of a transformed image?

Ex. No. HISTOGRAM PROCESSING AND BASIC Date

THRESHOLDING FUNCTIONS
AIM:
The aim of this project is to perform histogram processing and basic thresholding
functions on digital images to enhance or segment regions of interest based on their pixel
intensity values.

SOFTWARE REQUIRED:
MATLAB 2013b

THEORY:
A histogram is a graphical representation of the distribution of pixel intensity values in an
image. It provides insights into the image's contrast, brightness, and distribution of features.
Histogram processing refers to operations that modify the pixel intensity distribution in an
image. Common processes include contrast stretching, histogram equalization, and histogram
matching. Thresholding is a technique used to segment an image into foreground and
background regions based on a specific threshold value. Pixels with intensity values above the
threshold are classified as foreground, while those below are considered background.

PROCEDURE:
1. Load the digital image you want to process using MATLAB.
2. Compute and display the histogram of the image, representing the frequency of each
intensity level.
3. Implement contrast stretching by specifying a lower and upper bound for the pixel
intensity values. This expands the range of pixel values, enhancing image contrast.
4. Perform histogram equalization to redistribute pixel intensities across the entire range,
which improves the overall image contrast.
5. If needed, match the image's histogram to a predefined histogram to make it similar to a
reference image's histogram.
6. Apply a thresholding operation by specifying a threshold value. Pixels with values above
this threshold are set to a predefined value (e.g., 255 for white), while those below are set
to another predefined value (e.g., 0 for black).
7. Display the original image, histogram, and the processed image after histogram
processing and thresholding.

St. Joseph’s Institute of Technology 18

CS4509- Digital image processing Laboratory Department of ADS 2023-2024

PROGRAM:

%Creating histogram
clc;
clear
all;
close
all;
a = rgb2gray(imread('image.jpg'));
[m, n] = size(a);
no = 0:255;
count = 0;
for z=1:256
for i=1:m
for j=1:n
if a(i, j) == z-1;
count = count + 1;
end
end
end
t(z) = count;
count = 0;
end
figure;
imshow(a)
title('Original image')
figure;
stem(no, t)
grid on;
xlabel('Intesity levels')
ylabel('No. of pixels')
title('Histogram of the image')

%Histogram equalization
clc
clear
all
close
all
a = rgb2gray(imread('image.jpg'));
[m, n] = size(a);
no_of_pixels = m*n;
figure;
imshow(a)
title('Original image')
h_im = uint8(zeros(m, n));
count = zeros(256, 1);
probf = zeros(256, 1);
St. Joseph’s Institute of Technology 19
CS4509- Digital image processing Laboratory Department of ADS 2023-2024

probc = zeros(256, 1);

cdf = zeros(256, 1);
output = zeros(1, 256);
no = 1:256;
for i=1:m
for j=1:n
value = a(i, j);
count(value + 1) = count(value + 1) + 1;
probf(value + 1) = count(value +1/no_of_pixels;
end
end
figu
re;
stem(no, probf)
title('Probability distribution function')sum = 0;
b = 255;
for i=1:size(probf)
sum = sum + count(i);
cdf(i) = sum;
probc(i) = cdf(i)/no_of_pixels;
output(i) = round(probc(i) * b);
end
figu
re;
stem(no, output)
title('Rounded off CDF*(L-1)')
for i=1:m
for j=1:n
h_im(i, j) = output(a(i, j) + 1);
end
end
figure; imshow(h_im)
title('Histogram equalization')
% Thresholding the original image using Otsu's method
a = rgb2gray(imread('image.jpg'));
level = graythresh(a);
B = imbinarize(a, level);
subplot(1, 2, 1) imshow(uint8(a))
subplot(1, 2, 2) imshow(B)

St. Joseph’s Institute of Technology 20

CS4509- Digital image processing Laboratory Department of ADS 2023-2024

OUTPUT:

St. Joseph’s Institute of Technology 21

CS4509- Digital image processing Laboratory Department of ADS 2023-2024

RESULT:

REVIEW QUESTIONS:
1. What is a histogram in the context of digital image processing, and why is it useful for
analyzing images?
2. Explain the purpose of contrast stretching and histogram equalization in image
enhancement.

St. Joseph’s Institute of Technology 22

CS4509- Digital image processing Laboratory Department of ADS 2023-2024

3. What is the fundamental concept of thresholding, and how is it used to segment regions
in an image?
4. How can you determine an appropriate threshold value for thresholding an image?
5. What are the key factors to consider when assessing the quality of an image after
histogram processing and thresholding?

Ex. No. COMPUTATION OF MEAN, STANDARD Date

DEVIATION, CORRELATION COEFFICIENT OF
THE GIVEN IMAGE

AIM:
The aim of this experiment is to compute the Mean, Standard Deviation, and Correlation
Coefficient of a given image.

SOFTWARE REQUIRED:
MATLAB 2013b

THEORY:
The mean (average) of an image represents the central tendency of pixel intensities in the
image. It is calculated by summing up all the pixel values and dividing by the total number
of pixels.
Mean (µ) = Σ [I(x, y)] / N
where I(x, y) represents the intensity of the pixel at position (x, y), and N is the total number
of pixels in the image. The standard deviation measures the amount of variation or dispersion
in pixel intensities. It provides information about the image's contrast and the distribution of
pixel values. Standard Deviation (σ) = √[Σ [(I(x, y) - µ)²] / N]
where I(x, y) represents the pixel intensity, µ is the mean, and N is the total number of pixels
in the image.
The correlation coefficient measures the degree of linear relationship between two images.
It helps to understand how similar or dissimilar two images are. A high correlation coefficient
indicates strong positive correlation, while a low coefficient suggests little to no correlation.
Correlation Coefficient (ρ) = Σ [(I1(x, y) - µ1) * (I2(x, y) - µ2)] / [√Σ(I1(x, y) - µ1)² * √Σ(I2(x,
y) - µ2)²]
where I1(x, y) and I2(x, y) are the pixel intensities in two images, µ1 and µ2 are the means of
the respective images, and the summation is done over all pixels.

PROCEDURE:
1. Load the given image into your image processing environment MATLAB.
2. Calculate the Mean (µ) of the image using the formula mentioned above. This can be
done by iterating through all pixels and accumulating their values.
3. Compute the Standard Deviation (σ) of the image using the formula mentioned above.
Again, iterate through all pixels to calculate the necessary values.
4. If you need to compute the Correlation Coefficient between two images, repeat steps 2
and 3 for both images. Ensure that the images are of the same size and resolution.
5. Compute the Correlation Coefficient (ρ) using the formula for correlation mentioned
above.

St. Joseph’s Institute of Technology 23

CS4509- Digital image processing Laboratory Department of ADS 2023-2024

PROGRAM:
i=imread('cancercell.jpg');
subplot(2,2,1);
imshow(i);
title('Original Image');
g=rgb2gray(i);
subplot(2,2,2); imshow(g);
title('Gray Image');
c=imcrop(g);
subplot(2,2,3); imshow(c);
title('Cropped Image');
m=mean2(c);disp('m'); disp(m);
s=std2(c); disp('s'); disp(s);
figure, k=(checkerboard>0.8);
subplot(2,1,1); imshow(k);
title('Image1');
k1=(checkerboard>0.5);
subplot(2,1,2); imshow(k1);
title('Image2');
r=corr2(k,k1);
disp('r');disp(r);

OUTPUT:

St. Joseph’s Institute of Technology 24

CS4509- Digital image processing Laboratory Department of ADS 2023-2024

RESULT:

REVIEW QUESTIONS:
1. How is the mean computed for pixel values in an image and what does it represent?
2. What does the standard deviation reveal about an image's pixel values, and how is it
calculated?
3. What is the significance of the correlation coefficient in image analysis, and how is it
determined for two images?
4. Describe the concept of correlation coefficient as it pertains to image analysis. How is
the correlation coefficient calculated for two images, and what does it indicate about
their similarity or dissimilarity?
5. How can mean and standard deviation be used to assess the overall brightness and
contrast of an image? Provide a step-by-step explanation with a practical example.

Ex. No. IMPLEMENTATION OF IMAGE Date

ENHANCEMENT-SPATIAL FILTERING

AIM:
The aim of this project is to implement image enhancement using spatial filtering techniques
to improve the visual quality of digital images.

SOFTWARE REQUIRED:
MATLAB 2013b

THEORY:
Image enhancement through spatial filtering is a fundamental image processing technique. It
involves the use of convolution operations with specific filter kernels to modify pixel values in
an image. A basic spatial filter, such as an averaging filter, is used to reduce noise and enhance
St. Joseph’s Institute of Technology 25
CS4509- Digital image processing Laboratory Department of ADS 2023-2024

the overall appearance of the image. The process involves convolving the image with the filter
kernel, which replaces each pixel's value with a weighted average of its neighboring pixels.
This can help to improve image quality, reduce artifacts, and enhance specific features.

PROCEDURE:
1. Load the input image on which image enhancement will be performed.
2. Choose a suitable filter kernel based on the specific image enhancement goals.
Common choices include averaging filters, Gaussian filters, or edge-enhancing filters.
3. Convolve the input image with the selected filter kernel using the `conv2` function in
MATLAB.
4. Adjust the filter size and values to control the extent of enhancement.
5. Convert the resulting image to an appropriate data type (e.g., `uint8`) to ensure it falls
within the valid pixel value range (0-255).
6. Display both the original and enhanced images using MATLAB's `imshow` function.
7. Save the enhanced image to a file for further analysis or use.

PROGRAM:
% Load the image
originalImage = imread('image2.jpg'); % the image filename

% Convert the image to double precision for better processing

originalImage = im2double(originalImage);

% Define the filter kernel (averaging filter)

filterSize = 3; % Size of the filter kernel (e.g., 3x3)
filter = ones(filterSize) / filterSize^2;

% Apply the filter using convolution

enhancedImage = imfilter(originalImage, filter, 'replicate'); % 'replicate' padding to handle
image borders

% Display the original and enhanced images

subplot(1, 2, 1);
imshow(originalImage);
title('Original Image');
subplot(1, 2, 2);
imshow(enhancedImage);
title('Enhanced Image');

% Save the enhanced image if needed

% imwrite(enhancedImage, 'enhanced_image.jpg');

St. Joseph’s Institute of Technology 26

CS4509- Digital image processing Laboratory Department of ADS 2023-2024

OUTPUT:

RESULT:

REVIEW QUESTIONS:
1. What is the primary goal of image enhancement through spatial filtering?
2. Can you explain the concept of a filter kernel and its role in image enhancement using spatial
filtering?
3. What is convolution, and how does it apply to image enhancement in this context?
4. What are some common spatial filter types used for image enhancement, and how do they
differ in their effects on an image?
5. What considerations should be taken into account when selecting a filter kernel and its size
for image enhancement, and how does this choice affect the output?

Ex. No. IMPLEMENTATION OF IMAGE Date

ENHANCEMENT- FILTERING IN
FREQUENCY DOMAIN

AIM:
The aim of this project is to implement image enhancement through filtering in the
frequency domain to improve the visual quality and extract specific features from digital
images.
SOFTWARE REQUIRED:
MATLAB 2013b
THEORY:
Image enhancement through filtering in the frequency domain involves transforming an image
from the spatial domain to the frequency domain using techniques like the Fast Fourier
Transform (FFT). Once in the frequency domain, filtering operations are applied to enhance or
suppress specific frequency components. The filtered image is then transformed back to the
spatial domain using the Inverse Fast Fourier Transform (IFFT). This process can be used to
remove noise, sharpen details, and improve image quality.
St. Joseph’s Institute of Technology 27
CS4509- Digital image processing Laboratory Department of ADS 2023-2024

PROCEDURE:
1. Load the input image for image enhancement.
2. Apply the Fast Fourier Transform (FFT) to convert the image from the spatial domain
to the frequency domain.
3. Choose an appropriate filter (e.g., high-pass, low-pass, or band-pass filter) to perform
the desired enhancement operation.
4. Apply the selected filter to the transformed image in the frequency domain. This
involves point-wise multiplication or convolution with the filter.
5. Use the Inverse Fast Fourier Transform (IFFT) to transform the filtered image back to
the spatial domain.
6. Adjust the output image to ensure it falls within the valid pixel value range (0-255).
7. Display both the original and enhanced images.
PROGRAM:
clc
clear all
close all
a = imresize(imread('image.jpg’), [256, 256]);
m, n] = size(a);
mask = zeros(m, n);
for i=113:143
for j=113:143
mask(i, j) = 1;
end
end
b = fftshift(fft2(a));
d = b .* mask;
e = abs(ifft2(d));
subplot(2, 2, 1)
imshow(a)
title('Orginal image')
subplot(2, 2, 2) imshow(uint8(e))
title('Low-pass filter filtered image')
subplot(2, 2, 3)
imshow(mask)
title('Low-pass filter mask')

OUTPUT:

St. Joseph’s Institute of Technology 28

CS4509- Digital image processing Laboratory Department of ADS 2023-2024

RESULT:

REVIEW QUESTIONS:
1. What is the fundamental difference between image enhancement in the spatial domain and the
frequency domain?
2. How does the Fast Fourier Transform (FFT) enable image enhancement in the frequency
domain?
3. What are some common filters used for image enhancement in the frequency domain, and
how do they affect image features?
4. Can you explain the concept of convolution or point-wise multiplication in the context of
frequency domain filtering for image enhancement?
5. What are the advantages of using frequency domain filtering for image enhancement, and in
what scenarios is it particularly useful?

Ex. No. IMAGE SEGMENTATION – EDGE Date

DETECTION, LINE DETECTION AND
POINT DETECTION
AIM:
The aim of this project is to perform image segmentation, specifically focusing on edge
detection, line detection, and point detection in digital images, enabling the extraction of
important features and boundaries.

SOFTWARE REQUIRED:
MATLAB 2013b
THEORY:
Image segmentation is a crucial image processing task that aims to partition an image into
meaningful regions. The techniques used in this project involve:
1. Edge Detection: This technique highlights significant transitions or boundaries in an image.
It is commonly used to identify object boundaries or regions of interest.
2. Line Detection: Utilizing the Hough Transform, this technique identifies straight lines in an
image, which is useful for applications such as object tracking and lane detection.
3. Point Detection: The Harris corner detection method identifies key points or corners in an
image, providing critical feature points for various computer vision tasks.
PROCEDURE:
1. Load the input image onto which image segmentation techniques will be applied.
2. Apply edge detection methods, such as the Canny edge detector, to highlight image
edges and transitions.
3. Employ the Hough Transform to identify straight lines in the image, displaying their
positions and orientations.
4. Utilize the Harris corner detection technique to find key points or corners in the image.
5. Display the original image and the results of edge detection, line detection, and point
detection.
6. Save the images resulting from each segmentation technique for further analysis or use.
PROGRAM:
% Load the image
inputImage = imread('your_image.jpg');
% Edge Detection using Canny
edgeImage = edge(rgb2gray(inputImage), 'Canny');
% Line Detection using Hough Transform
St. Joseph’s Institute of Technology 29
CS4509- Digital image processing Laboratory Department of ADS 2023-2024

[H, theta, rho] = hough(edgeImage);

peaks = houghpeaks(H, 10); % Adjust the number of peaks as needed
lines = houghlines(edgeImage, theta, rho, peaks);
% Point Detection using the Harris corner detector
corners = detectHarrisFeatures(rgb2gray(inputImage));
% Display the results
figure;
subplot(2, 2, 1);
imshow(inputImage);
title('Original Image');
subplot(2, 2, 2);
imshow(edgeImage);
title('Edge Detection (Canny)');
subplot(2, 2, 3);
imshow(inputImage);
hold on;
for k = 1:length(lines)
xy = [lines(k).point1; lines(k).point2];
plot(xy(:,1), xy(:,2), 'LineWidth', 2, 'Color', 'red');
end
title('Line Detection (Hough Transform)');
subplot(2, 2, 4);
imshow(inputImage);
hold on;
plot(corners.selectStrongest(50));
title('Point Detection (Harris Corners)');
% Save the images if needed
imwrite(edgeImage, 'edge_detection.jpg');
imwrite(inputImage, 'line_detection.jpg');
imwrite(inputImage, 'point_detection.jpg');
OUTPUT:

St. Joseph’s Institute of Technology 30

CS4509- Digital image processing Laboratory Department of ADS 2023-2024

RESULT:

REVIEW QUESTIONS:
1. What is the significance of edge detection in image segmentation?
2. How does the Hough Transform assist in line detection in image analysis?
3. What are the key objectives of point detection using methods like the Harris corner detector?
4. In what scenarios might image segmentation techniques like edge detection, line detection,
and point detection be particularly valuable?
5. Can you discuss potential challenges or limitations associated with these image segmentation
techniques?

Ex. No. IMPLEMENTATION OF REGION BASED Date

SEGMENTATION
AIM:
The aim is to partition an input image into meaningful regions or segments based on color,
intensity, or texture characteristics, thus improving image analysis and object recognition.

SOFTWARE REQUIRED:
MATLAB 2013b

THEORY:
Region-based segmentation is a computer vision technique that groups pixels or image regions
into meaningful segments based on their similarity in terms of color, intensity, or texture. This
segmentation technique involves the following key steps:
1. Preprocessing: The input image is preprocessed to enhance its quality, reduce noise, and
normalize intensity values.
2. Feature Extraction: Relevant features, such as color histograms, texture features, or intensity
gradients, are computed for each pixel or region in the image.
3. Region Growing or Split-and-Merge: The segmentation algorithm selects seed points and
iteratively grows or splits regions based on feature similarity. Region growing starts with seed
pixels and expands by including neighboring pixels that meet certain similarity criteria. Split-
and-merge, on the other hand, recursively divides large regions into smaller ones if
dissimilarity criteria are met.
4. Region Merging: After regions have been created, regions that are too small or similar to
their neighbors are merged.
5. Postprocessing: Additional postprocessing steps can be applied to refine the segmentation
results, such as removing small noise regions or smoothing region boundaries.
PROCEDURE:
1. Import the input image.
2. Preprocess the image (e.g., noise reduction, intensity normalization).
3. Compute relevant image features (e.g., color histograms, texture features).
4. Initialize seed points or regions.
5. Implement the region growing or split-and-merge algorithm.
6. Apply region merging to refine the segmentation.
7. Perform postprocessing as needed.
8. Visualize and analyze the segmented image.
St. Joseph’s Institute of Technology 31
CS4509- Digital image processing Laboratory Department of ADS 2023-2024

9. Optimize parameters and iterate for better results.

10. Save or export the segmented image.
PROGRAM:
% Read the input image
inputImage = imread('clock.jpg');
% Define a seed point (you can choose this point manually)
seedPoint = [100, 100]; % [row, column]
% Set the intensity similarity threshold
intensityThreshold = 20;
% Initialize the segmented image
segmentedImage = zeros(size(inputImage));
% Create a stack for region growing
stack = [];
stack = [stack; seedPoint];
% Define the region mean and region size
regionMean = double(inputImage(seedPoint(1), seedPoint(2)));
regionSize = 1;
% Define connectivity for 8-connected neighbors
connectivity = [-1, -1; -1, 0; -1, 1; 0, -1; 0, 1; 1, -1; 1, 0; 1, 1];
% Region growing loop
while ~isempty(stack)
currentPoint = stack(1, :);
stack(1, :) = [];
for i = 1:8
neighbor = currentPoint + connectivity(i, :);
if neighbor(1) > 0 && neighbor(1) <= size(inputImage, 1) && neighbor(2) > 0 &&
neighbor(2) <= size(inputImage, 2)
if segmentedImage(neighbor(1), neighbor(2)) == 0 &&
abs(double(inputImage(neighbor(1), neighbor(2))) - regionMean) <= intensityThreshold
regionSize = regionSize + 1;
regionMean = (regionMean * (regionSize - 1) + double(inputImage(neighbor(1),
neighbor(2)))) / regionSize;
stack = [stack; neighbor];
segmentedImage(neighbor(1), neighbor(2)) = 1;
end
end
end
end
% Display the segmented image
figure;
subplot(1, 2, 1);
imshow(inputImage);
title('Original Image');
subplot(1, 2, 2);
imshow(segmentedImage, []);
title('Segmented Image');
% You can save the segmented image if needed
imwrite(segmentedImage, 'segmented_image.jpg');

St. Joseph’s Institute of Technology 32

CS4509- Digital image processing Laboratory Department of ADS 2023-2024

OUTPUT:

RESULT:

REVIEW QUESTIONS:
1. What is the primary objective of region-based image segmentation?
2. Describe the key steps involved in region-based segmentation.
3. What are some preprocessing techniques that can be applied to improve segmentation
results?
4. How does region growing differ from split-and-merge in image segmentation?
5. What postprocessing steps can be employed to refine the segmented image?

Ex. No. Date

BASIC MORPHOLOGICAL OPERATIONS
AIM:
The aim is to understand and implement basic morphological operations, including dilation,
erosion, opening, and closing, using digital images.

SOFTWARE REQUIRED:
MATLAB 2013b

THEORY:
Morphological operations are image processing techniques based on the shape and structure of
objects within an image. They are often applied to binary or grayscale images and rely on a
structuring element (also known as a kernel) to modify the pixels in the image.
Dilation is used to enlarge the boundaries of objects in a binary image. It involves sliding the
structuring element over the image and replacing the center pixel with the maximum value
found in the neighborhood defined by the structuring element.
Erosion is used to shrink the boundaries of objects in a binary image. It involves sliding the
structuring element over the image and replacing the center pixel with the minimum value
found in the neighborhood defined by the structuring element.
Opening is a combination of erosion followed by dilation. It is used to remove small noise in
binary images or separate objects that are close to each other.
St. Joseph’s Institute of Technology 33
CS4509- Digital image processing Laboratory Department of ADS 2023-2024

Closing is a combination of dilation followed by erosion. It is used to close small gaps in binary
images or to connect objects that are almost touching.

PROCEDURE:
1. Import the digital image onto which you want to apply morphological operations.
2. Convert the image to a binary format if it's not already binary. This is often done by
thresholding.
3. Define a structuring element (a small matrix) that specifies the neighborhood for the
morphological operation. The size and shape of the structuring element depend on the
specific operation and the desired effect.
4. Apply dilation to the binary image using the chosen structuring element.
5. Update the pixel values according to the dilation operation.
6. Apply erosion to the binary image using the chosen structuring element.
7. Update the pixel values according to the erosion operation.
8. Apply an opening operation, which consists of an erosion followed by dilation. This is
used for noise removal and object separation.
9. Apply a closing operation, which consists of a dilation followed by erosion. This is used
for gap-filling and object connection.
10. Visualize or save the processed image.
11. Experiment with different structuring elements and parameters to achieve the desired
image enhancement or feature extraction.
12. Summarize the results and the impact of the morphological operations on the image.

PROGRAM:
clc
clear all
close all
I = rgb2gray(imread('image.jpg'));
subplot(2, 3, 1)
imshow(I)
title('Original image')
%Edge detection and binary image
[~, threshold] = edge(I, 'sobel');
fudgeFactor = 0.5;
BWs = edge(I, 'sobel', threshold *fudgeFactor);
%Display the resulting binary gradient mask
subplot(2, 3, 2)
imshow(BWs)
title('Binary Gradient Mask')
%imdilate
se90 = strel('line', 3, 90);
se0 = strel('line', 3, 0);
BWsdil = imdilate(BWs, [se90, se0]);
subplot(2, 3, 3)
imshow(BWsdil)
title('Dilated Gradient Mask')
St. Joseph’s Institute of Technology 34
CS4509- Digital image processing Laboratory Department of ADS 2023-2024

BWdfill = imfill(BWsdil, 'holes');

subplot(2, 3, 4)
imshow(BWdfill)
title('Binary image with filled holes')
BWnobord = imclearborder(BWdfill, 4);
subplot(2, 3, 5)
imshow(BWnobord)
title('Cleared Border Image')
seD = strel('diamond', 1);
BWfinal = imerode(BWnobord, seD);
BWfinal = imerode(BWfinal, seD);
subplot(2, 3, 6)
imshow(BWfinal)
title('Segmented Image (imerode)')
%Remove inner elements
BW2 = bwmorph(BWfinal, 'remove');
figure;
subplot(2, 2, 1)
imshow(BW2)
title('Image with outline')
%Skeleton image
BW3 = bwmorph(BWfinal, 'skel', Inf);
subplot(2, 2, 2)
imshow(BW3)
title('Skeleton')
subplot(2, 2, 3) imshow(labeloverlay(I,
BWfinal)) title('Mask over original image')
BWoutline = bwperim(BWfinal); Segout = I;
Segout(BWoutline) = 255;
subplot(2, 2, 4)
imshow(Segout)
title('Outlined original image')
%Opening and closing
originalBW = I; figure;
subplot(2, 2, 1)
imshow(originalBW)
title('Original')
se = strel('disk', 10);
closeBW = imclose(originalBW, se);
subplot(2, 2, 2)
imshow(closeBW, [])
title('imclose')
original = I;
subplot(2, 2, 3)
imshow(original)
title('Original')
se = strel('disk', 20);
afterOpening = imopen(original, se);
subplot(2, 2, 4)
St. Joseph’s Institute of Technology 35
CS4509- Digital image processing Laboratory Department of ADS 2023-2024

imshow(afterOpening, [])
title('imopen')

OUTPUT:

St. Joseph’s Institute of Technology 36

CS4509- Digital image processing Laboratory Department of ADS 2023-2024

RESULT:

REVIEW QUESTIONS:
1. What are morphological operations in image processing, and what is their primary purpose?
2. Explain the fundamental morphological operations: dilation and erosion. How do they work,
and what are their typical applications?
3. What is the role of a structuring element in morphological operations? How does its size and
shape affect the results?
4. Describe the differences between opening and closing operations. When would you use one
over the other?
5. How can morphological operations be used for noise reduction or feature enhancement in
images? Provide examples of scenarios where morphological operations are beneficial.

Ex. No. IMPLEMENTATION OF IMAGE Date

COMPRESSION TECHNIQUES
AIM:
To perform JPEG compression using DCT to an image usingMATLAB.

SOFTWARE REQUIRED:
MATLAB 2013b

THEORY:
JPEG (Joint Photographic Experts Group) compression is a widely used image compression
technique that employs the Discrete Cosine Transform (DCT) for reducing the size of digital
images while maintaining reasonable image quality. JPEG compression begins with dividing
the image into small blocks, typically 8x8 pixels. These blocks are then subjected to the DCT
transformation, which converts spatial domain pixel values into frequency domain coefficients.
After the DCT, quantization is applied to the frequency domain coefficients. Quantization
reduces the precision of the coefficients by dividing them by a quantization matrix. The
quantized coefficients are scanned in a zigzag pattern to group the high-frequency components
together and make them more compressible. The zigzag-scanned coefficients are then run-
length encoded, which replaces sequences of repeated values with a code that represents the
value and the number of repetitions. This is a form of lossless compression. The encoded data
is further compressed using Huffman encoding, which assigns shorter codes to more frequently
occurring sequences and longer codes to less frequent sequences.

PROCEDURE:
1. Load the input image.
2. Divide the image into 8x8 pixel blocks.
3. Apply the DCT transformation to each block.
4. Quantize the DCT coefficients using a quantization matrix.
5. Perform zigzag scanning on the quantized coefficients.
6. Run-length encode the zigzag-scanned coefficients.
7. Apply Huffman encoding to further compress the data.
8. Save the compressed data and encoding tables for later decoding.
9. To decompress, reverse the process using Huffman decoding, run-length decoding, de-
zigzag scanning, de-quantization, and inverse DCT transformation.
10. Display or save the decompressed image.
St. Joseph’s Institute of Technology 37
CS4509- Digital image processing Laboratory Department of ADS 2023-2024

PROGRAM:
%nearer values are retained apart from DC component
clc
clear
all
close
all
a = rgb2gray(imread('image.jpg'));
figure;
subplot(1, 2, 1)
imshow(a)
title('Original image')
[m, n] = size(a);
bs = 8;
Q = 20 * ones(bs);
Q(1, 1) = 10;
Rec_a = zeros(size(a));
for i=1:bs:(m-bs+1)
for j=1:bs:(n-bs+1)
block = a(i:i+bs-1, j:j+bs-1);
block_D = dct2(block);
block_Q = round(block_D./Q);
d = block_Q.*Q;
Rec_a(i:i+bs-1, j:j+bs-1) = idct2(d);
end
end
subplot(1, 2, 2)
imshow(uint8(Rec_a))
title('Reconstructed image')
MSE = sum(sum((uint8(a)-uint8(Rec_a)).^2))/(m*n)
% set all components to zero except the DC component
clc
clear
all
close
all
a = rgb2gray(imread('D:\College\DIP\Lab\Lab8\image.jpg'));
figure;
subplot(1, 2, 1)
imshow(a)
title('Original image')
[m, n] = size(a);
bs = 8;
Q = 20 * ones(bs);
Q(1, 1) = 10;
Rec_a = zeros(size(a));
for i=1:bs:(m-bs+1)
for j=1:bs:(n-bs+1)
block = a(i:i+bs-1, j:j+bs-1);
St. Joseph’s Institute of Technology 38
CS4509- Digital image processing Laboratory Department of ADS 2023-2024

block_D = dct2(block);
d = zeros(size(block_D));
d(1, 1) = block_D(1, 1);
Rec_a(i:i+bs-1, j:j+bs-1) = idct2(d);
end
end
subplot(1, 2, 2)
imshow(uint8(Rec_a))
title('Reconstructed image')
MSE = sum(sum((uint8(a)-uint8(Rec_a)).^2))/(m*n)
% specify how many elements are to be retained
clc
clear all
close all
a = rgb2gray(imread('D:\College\DIP\Lab\Lab8\image.jpg'));
k = input("Enter the number of values to be retained:");
figure; subplot(1,
2, 1)imshow(a)
title('Original image');
[m, n]=size(a);
bs = 8;
Rec_a = zeros(size(a));for
i=1:bs:(m-bs+1)
for j=1:bs:(n-bs+1)
block = a(i:i+bs-1, j:j+bs-1);
block_D = dct2(block);
d = zeros(size(block_D));
d(1:k, 1:k) = block_D(1:k, 1:k);
Rec_a(i:i+bs-1, j:j+bs-1) = idct2(d);
end
end
subplot(1, 2, 2)
imshow(uint8(Rec_a))
title('Reconstructed image');
MSE = sum(sum((uint8(a)-uint8(Rec_a)).^2))/(m*n)
OUTPUT:

Fig 1.
St. Joseph’s Institute of Technology 39
CS4509- Digital image processing Laboratory Department of ADS 2023-2024

Fig.2.

Fig.3.

Enter the number of values to be retained: 5

MSE = 4.9197
RESULT:

REVIEW QUESTIONS:
1. What is the fundamental purpose of JPEG compression, and how does it achieve this
purpose?
2. How does the Discrete Cosine Transform (DCT) contribute to JPEG compression?
3. Explain quantization in JPEG compression. How does the choice of quantization matrix
affect image quality?
4. Describe the significance of zigzag scanning in JPEG compression. How does it impact the
encoding process?
5. Why are both Run-Length Encoding (RLE) and Huffman Encoding used in JPEG
compression?

St. Joseph’s Institute of Technology 40

CS4509- Digital image processing Laboratory Department of ADS 2023-2024

S.No Prepared by Sign

1 Dr.P.G.V.Ramesh

S.No CO,PO,PSO,PEO Sign

verified by
1 Dr.Jeba Johannah Samuel

Dr.C.Mohan

Mr.J.Bino

S.No Subject Domain- Staff Sign

Members
1 Dr.C.Gnana Kouslaya

Mr.T.Siva

Mrs.Maria Cristina Blessy

Ms.Sindhu kavi

St. Joseph’s Institute of Technology 41