How to do 2D Image Processing?

Example

[Joshua]

%% =========================

% [3.1] Load and display image

% =========================

imOriginal = imread("Section3.png");

figure;

imshow(imOriginal);

title("Original MRI Image");

% If image is RGB, convert to grayscale (usually this file is already grayscale)

if ndims(imOriginal) == 3

imOriginal = rgb2gray(imOriginal);

end

%% =========================

% [3.1] Remove salt-and-pepper noise using median filter

% =========================

imMed = medfilt2(imOriginal); % 2D median filter

figure;

imshowpair(imOriginal, imMed, "montage", "scaling", "none");

title("Original (left) vs Median Filtered (right)");

%% =========================

% [3.1] Compare with averaging filter (why median works better for S&P noise)

% =========================

avgFilt = (1/9) * ones(3,3);

% filter2 returns double, so cast back for display if needed

imAvg = filter2(avgFilt, imOriginal);

imAvg = uint8(imAvg);

figure;

imshowpair(imOriginal, imAvg, "montage", "scaling", "none");

title("Original (left) vs Averaging Filter (right)");

%% =========================

% [3.1] Histogram of median-filtered image

% Lab says use median-filtered output for intensity analysis

% =========================

figure;

imhist(imMed);

title("Histogram of Median-Filtered Image");

%% =========================

% [3.1] Create logical mask for NON-background pixels

% Black background pixel value = 0 (for uint8 grayscale)

% =========================

logicArr = (imMed ~= 0); % 1 = non-background, 0 = black background

%% =========================

% [3.1] Contrast adjustment using ONLY non-background pixels

% Steps:

% 1) extract non-background pixels

% 2) compute intensity limits from those pixels

% 3) use imadjust on the whole image with those limits

% =========================

nonBgPixels = imMed(logicArr);

% Simple version (directly from image values)

inLow = double(min(nonBgPixels)) / 255;

inHigh = double(max(nonBgPixels)) / 255;

% Optional better version (more robust to outliers):

% pr = prctile(nonBgPixels, [1 99]);

% inLow = double(pr(1))/255;

% inHigh = double(pr(2))/255;

% Contrast-adjust image

imAdj = imadjust(imMed, [inLow inHigh], []);

%% Histogram after adjustment (should use more of the intensity range)

figure;

imhist(imAdj);

title("Histogram After Contrast Adjustment");

%% =========================

% [3.1] Final comparison: original vs median-filtered + contrast-adjusted

% =========================

figure;

imshowpair(imOriginal, imAdj, "montage", "scaling", "none");

title("Original (left) vs Median Filtered + Contrast Adjusted (right)");

[Mathieu NOE]

again, what is your question ?

Answer 1

[Image Analyst]

For some examples, see my File Exchange:

https://www.mathworks.com/matlabcentral/fileexchange/?term=authorid%3A31862&sort=downloads_desc

especially the Image Segmentation Tutorial submission.

And see my attached salt and pepper noise removal demos.