How to normalize heatmap image and find the shape info

The following heatmap image is non-normalized and includes RCS info. How to normalize it and use the shape of the object for classification instead of RCS?

Answer 1

[Darshak]

Hi Mohamed,

When working with heatmaps for data analysis, I've had similar experiences where MATLAB made the process much more manageable. Below, I've provided a sample code that can be used for the task. Feel free to adapt its individual components for other needs as well.

heatmapImage = imread('your_heatmap_image.png');
if size(heatmapImage, 3) == 3
    heatmapImage = rgb2gray(heatmapImage);
end
normalizedHeatmap = normalize(double(heatmapImage));
figure;
imshow(normalizedHeatmap, []);
title('Normalized Heatmap');
binaryImage = imbinarize(normalizedHeatmap);
stats = regionprops(binaryImage, 'Area', 'Perimeter', 'BoundingBox', 'Centroid');
figure;
imshow(binaryImage);
hold on;
for k = 1:length(stats)
    rectangle('Position', stats(k).BoundingBox, 'EdgeColor', 'r', 'LineWidth', 1);
end
title('Extracted Shapes');
hold off;

The result for the given image is as follows:

There are multiple functions used here, you can explore the same using the following commands:

• “doc normalize”
• “doc imbinarize”
• “doc regionprops”

I hope this helps with your issue.