How to check that an image having bimodal or multimodal distribution in histogram?

If I am having some gray scale input images then how from the histogram plots we can analyze that the histogram having bimodal or multimodal histogram? If the histogram is multimodal then is this possible to mark the threshold value using any red line or some other things automatically?

Answer 1

[Jaswanth]

Hi,

To determine if a grayscale image's histogram is bimodal or multimodal using MATLAB, start by loading grayscale image and calculating its histogram to understand the distribution of pixel intensities.

img = imread('your_image.png'); % Load the grayscale image
histogramValues = imhist(img); % Calculate the histogram

Plot the histogram to visually assess the distribution.

figure;
bar(histogramValues);

To detect the presence of multiple peaks, which indicate bimodal or multimodal distributions, you can use MATLAB's “findpeaks” function.

[pks, locs] = findpeaks(histogramValues);

Evaluate the number of peaks to determine the histogram's nature.

numPeaks = numel(pks);
if numPeaks >= 2
    disp('The histogram is bimodal or multimodal.');
else
    disp('The histogram is unimodal.');
end

If the histogram is multimodal, you can automatically highlight potential threshold values by marking valleys between peaks.

[valleys, valleyLocs] = findpeaks(-histogramValues);
valleyLocs = valleyLocs(valleys < 0); % Identify valley locations

figure;
bar(histogramValues);
hold on;
plot(locs, pks, 'ro', 'MarkerFaceColor', 'r'); % Highlight peaks
for i = 1:length(valleyLocs)
    xline(valleyLocs(i), 'r', 'LineWidth', 2); % Mark valleys as thresholds
end
hold off;

You may refer to the following MathWorks documentation to know more about the functions mentioned above:

• imhist: https://www.mathworks.com/help/images/ref/imhist.html
• bar: https://www.mathworks.com/help/matlab/ref/bar.html
• findpeaks: https://www.mathworks.com/help/signal/ref/findpeaks.html

I hope the solution provided above is helpful.