Myujth

I have a task regarding the matlab image processing toolbox. The task is the following:

My solution to these steps is:

I = imread('Ball.jpg');

I1 = imnoise(I, 'salt&pepper', 0.2);

G = rgb2gray(I1);

C = fspecial('Laplacian',h);

imwrite(C, 'clean.jpg');

subplot(1,2,1);
imshow(I1,);

subplot(1,2,2);
imshow(C,);

I think you made quite some mistakes.

First, the image you read is already noisy, as it doesn't specifically say "add noise to the image". This makes your second step, imnoise, redundant.

Second, by using fspecialyou are creating a filter. In this case its type is a Laplacian filter of a given alpha alpha (between 0 and 1). That alone doesn't filter your image. You have to use the function imfilter in order to process the image.

I = imread('Ball.jpg');

G = rgb2gray(I);

h = fspecial('Laplacian',0.7); % 0.1 is the alpha, try out which one suits your case the most

C = imfilter(G,h);

imwrite(C, 'clean.jpg');

subplot(1,2,1);
imshow(I,);

subplot(1,2,2);
imshow(C,);

Note, that the Laplacian filter doesn't have to be the most suitable for you. There are lots of filter types listed in the MatLab documentation that you can use. Consider using a Gaussian filter.

Your solution is incomplete, for example you don't apply your filter on your noisy picture. Here is an example that might work :

%% Load image (I.)
I = imread('Ball.jpg');

%% Convert image into grayscale (II.)
G = rgb2gray(I);

%% Add noise (if 'Ball.jpg' isn't already noisy)
I1 = imnoise(G, 'salt & pepper', 0.2); % NB : imnoise needs the image to be grayscale

%% Create the filter (III.)
C = fspecial('Laplacian');

%% Apply the filter (III.)
IClean = filter2(C,I1);

%% Write the picture in new file (IV.)
imwrite(IClean, 'clean.jpg');

%% Display images (V.)
subplot(1,2,1), imshow(I1,);
subplot(1,2,2), imshow(IClean,);

Depending on the result, you can validate your idea of an "appropriate spatial domain filter" in question III.