Question

2. (20 marks) Given a noised color image test.jpg', first use media filter to remove noise from R, G, B channels respectively, then enhancement each channel by using histogram equalization, then combine them and use Roberts operator to detect its edges. Finally, display the original image 'test.jpg', the brighter image, and its edges figure. in one

Answer 1

% This is the matlab code for the question asked

% save the input image as 'test.jpg' in the same folder where the code is saved

clear all; close all; clc;
test = imread('test.jpg');
imshow(test);
title('Original Test Image')
%% median filtering each channels
outR = medfilt2(test(:,:,1));
outG = medfilt2(test(:,:,2));
outB = medfilt2(test(:,:,3));
%combining channels together
out1 = zeros(size(test),'uint8');
out1(:,:,1) = outR;
out1(:,:,2) = outG;
out1(:,:,3) = outB;

figure, imshow(out1);
title('Image after median filtering');
%% Histogram Equalization of individual channels
outR = histeq(outR);
outG = histeq(outG);
outB = histeq(outB);
%combining different channels together
out2 = zeros(size(out1),'uint8');
out2(:,:,1) = outR;
out2(:,:,2) = outG;
out2(:,:,3) = outB;
figure, imshow(out2);
title('Image after histogram equalization');
%% Robert's operator for edge detection
edgeR = edge(outR,'Roberts');
edgeG = edge(outG,'Roberts');
edgeB = edge(outB,'Roberts');
% ORing all three channels, since output of edge are logical
EdgeImg = edgeR + edgeB + edgeG;
figure, imshow(EdgeImg);
title('Image after applying Roberts operator for edge detection');