Regional Descriptors in Image Description for DIP and its implementation in MATLAB || Description||

Video lecture series on Digital Image Processing, Lecture: 71,
Regional Descriptors in Image Description for DIP and its implementation in MATLAB || Representation and Description||
What are simple regional descriptors?
What are topological descriptors?
How to calculate Euler number?
What are Textures?
Implementation of Regional descriptors in MATLAB
Link to download ppts/lecture notes:
drive.google.com/drive/folder...
#DIP
#DIPwithMATLAB
#DigitalImageProcessingUsingMATLAB
#digitalimageprocessing
#digitalimageprocessinglecturesin4K
#studywithdrdafda
The MATLAB code used in the video is given below:
% MATLAB program for Simple Regional descriptors
binary_image = zeros(100, 100);
[x, y] = meshgrid(1:100, 1:100);
circle = (x - 50).^2 + (y - 50).^2 (less than or equal to) 25^2;
binary_image(circle) = 1; % Set circle region to 1 (white)
% Calculate region properties
stats = regionprops(binary_image, 'Area', 'Perimeter', 'BoundingBox');
% Calculate Compactness, Circularity Ratio, Rectangularity, and Elongatedness
area = stats.Area;
perimeter = stats.Perimeter;
boundingBoxArea = prod(stats.BoundingBox(3:4));
circularity = (4 * pi * area) / (perimeter^2);
circularityRatio = area / boundingBoxArea;
rectangularity = area / boundingBoxArea;
elongatedness = max(stats.BoundingBox(3:4)) / min(stats.BoundingBox(3:4));
% Display the binary image with region properties
imshow(binary_image);
title('Binary Image with Region Properties');
% Display calculated descriptors in the command window
disp(['Area: ', num2str(area)]);
disp(['Perimeter: ', num2str(perimeter)]);
disp(['Compactness: ', num2str(4 * pi)]);
disp(['Circularity Ratio: ', num2str(circularityRatio)]);
disp(['Rectangularity: ', num2str(rectangularity)]);
disp(['Elongatedness: ', num2str(elongatedness)]);
% MATLAB program for Regional descriptors (Euler number calculation)
% Create binary images with holes
binary_image_one_hole = ones(100, 100); % Initialize with ones (white)
binary_image_two_holes = ones(100, 100);
binary_image_three_holes = ones(100, 100);
% Create a hole
[x1, y1] = meshgrid(1:100, 1:100);
circle1 = (x1 - 50).^2 + (y1 - 50).^2 (less than or equal to) 25^2;
binary_image_one_hole(circle1) = 0; % Set circle region to 0 (black)
% Create two holes
circle2 = (x1 - 30).^2 + (y1 - 30).^2 (less than or equal to) 10^2;
circle3 = (x1 - 70).^2 + (y1 - 70).^2 (less than or equal to) 10^2;
binary_image_two_holes(circle2) = 0; % Set circle regions to 0 (black)
binary_image_two_holes(circle3) = 0;
% Create three holes
circle4 = (x1 - 25).^2 + (y1 - 75).^2 (less than or equal to)10^2;
circle5 = (x1 - 50).^2 + (y1 - 50).^2 (less than or equal to) 10^2;
circle6 = (x1 - 75).^2 + (y1 - 25).^2 (less than or equal to) 10^2;
binary_image_three_holes(circle4) = 0; % Set circle regions to 0 (black)
binary_image_three_holes(circle5) = 0;
binary_image_three_holes(circle6) = 0;
euler_number_one_hole = bweuler(binary_image_one_hole);
euler_number_two_holes = bweuler(binary_image_two_holes);
euler_number_three_holes = bweuler(binary_image_three_holes);
subplot(1, 3, 1);
imshow(binary_image_one_hole);
title(['Binary Image with 1 Hole (Euler Number: ', num2str(euler_number_one_hole), ')']);
subplot(1, 3, 2);
imshow(binary_image_two_holes);
title(['Binary Image with 2 Holes (Euler Number: ', num2str(euler_number_two_holes), ')']);
subplot(1, 3, 3);
imshow(binary_image_three_holes);
title(['Binary Image with 3 Holes (Euler Number: ', num2str(euler_number_three_holes), ')']);
disp(['Euler Number (1 Hole): ', num2str(euler_number_one_hole)]);
disp(['Euler Number (2 Holes): ', num2str(euler_number_two_holes)]);
disp(['Euler Number (3 Holes): ', num2str(euler_number_three_holes)]);
% MATLAB program for Regional descriptors (Textures)
%smooth texture
smooth_texture = imbinarize(rand(100, 100), 0.5);
% coarse texture (checkerboard pattern)
coarse_texture = checkerboard(10, 5) (greater than) 0.5;
regular_texture = repmat([zeros(1, 50), ones(1, 50)], 10, 1);
% Compute texture descriptors
co_occurrence_matrix = graycomatrix(regular_texture, 'Offset', [0 1]);
% (2) Structural texture descriptors
% - Apply edge detection
edges = edge(coarse_texture, 'Sobel');
% (3) Spectral texture descriptors
fourier_transform = fft2(smooth_texture);
% Display the generated textures
subplot(2, 4, 1);
imshow(smooth_texture);
title('Smooth Texture');
subplot(2, 4, 2);
imshow(coarse_texture);
title('Coarse Texture');
subplot(2, 4, 3);
imshow(regular_texture);
title('Regular Texture');
subplot(2, 4, 5);
imshow(co_occurrence_matrix, []);
title('Co-Occurrence Matrix');
descriptor1 = mean(co_occurrence_matrix(:));
subplot(2, 4, 6);
imshow(edges);
title('Edges');
descriptor2 = sum(edges(:));
subplot(2, 4, 7);
imshow(log(abs(fourier_transform)), []);
title('Fourier Transform (Log Magnitude)');
% Additional subplot for better visualization
subplot(2, 4, 8);
imshow(angle(fourier_transform), []);
title('Fourier Transform (Phase)');

[@user-yg6zy4pz2e]

Very helpful ❤

[@StudywithDrDafda]

% MATLAB program for Regional descriptors (Textures)
clc;
close all;
clear all;
% Create a binary image with a smooth texture
smooth_texture = imbinarize(rand(100, 100), 0.5);
% Create a binary image with a coarse texture (checkerboard pattern)
coarse_texture = checkerboard(10, 5) > 0.5;
% Create a binary image with a regular texture (stripes)
regular_texture = repmat([zeros(1, 50), ones(1, 50)], 10, 1);
% Compute texture descriptors
% (1) Statistical texture descriptors
% - Compute co-occurrence matrix
co_occurrence_matrix = graycomatrix(regular_texture, 'Offset', [0 1]);
% (2) Structural texture descriptors
% - Apply edge detection
edges = edge(coarse_texture, 'Sobel');
% (3) Spectral texture descriptors
% - Apply Fourier transform
fourier_transform = fft2(smooth_texture);
% Display the generated textures
subplot(2, 4, 1);
imshow(smooth_texture);
title('Smooth Texture');
subplot(2, 4, 2);
imshow(coarse_texture);
title('Coarse Texture');
subplot(2, 4, 3);
imshow(regular_texture);
title('Regular Texture');
% Display the computed descriptors
subplot(2, 4, 5);
imshow(co_occurrence_matrix, []);
title('Co-Occurrence Matrix');
descriptor1 = mean(co_occurrence_matrix(:));
subplot(2, 4, 6);
imshow(edges);
title('Edges');
descriptor2 = sum(edges(:));
subplot(2, 4, 7);
imshow(log(abs(fourier_transform)), []);
title('Fourier Transform (Log Magnitude)');
% Additional subplot for better visualization
subplot(2, 4, 8);
imshow(angle(fourier_transform), []);
title('Fourier Transform (Phase)');
% Adjust figure properties
colormap('gray');

[@StudywithDrDafda]

% MATLAB program for Regional descriptors (Euler number calculation)
clc;
close all;
clear all;
% Create binary images with holes
binary_image_one_hole = ones(100, 100); % Initialize with ones (white)
binary_image_two_holes = ones(100, 100);
binary_image_three_holes = ones(100, 100);
% Create a white circle (hole) in the first binary image
[x1, y1] = meshgrid(1:100, 1:100);
circle1 = (x1 - 50).^2 + (y1 - 50).^2

[@StudywithDrDafda]

% MATLAB program for Simple Regional descriptors
clc;
close all;
clear all;
% Create a binary image with a shape (white region on a black background)
binary_image = zeros(100, 100);
[x, y] = meshgrid(1:100, 1:100);
circle = (x - 50).^2 + (y - 50).^2