Extract the INNER white region of an image with black and white intersection.

Hello,

I would like to get the inner part of the white region in the image below, the region I want is the red rectangluar part (INNER part of the white region). I do not want the outter part of the white region.

However, with my code below, I can only get the out part of the white region, is there anyone can help me?

Thanks a lot.

%%%%%My code
I = imread(filename);
Igray = rgb2gray(I);
BW2 = Igray < 50;
% Detect rectangle region (inside the line)
BW2 = imclearborder(~BW2);
BW2 = bwareafilt(BW2,1);
% Calculate bounding box
props2= regionprops(BW2,'BoundingBox');
boundingBox2 = props2.BoundingBox;
figure
imshow(BW2)

Answer 1

[Image Analyst]

First of all you don't need imclearborder because there is no part of your blob that touches the border. And after bwareafilt() you're probably still getting the original blob because some of the outer blobs are touching your inner blob by having the corners of the pixels touch. The default for bwareafilt() is 8-connected, which considers blobs touching on the diagonal to be the same blob. You should specify the "4-connected" option so that a blob is only connected if it's fully touching on the top, bottom, left, or right (not the diagonal/corner).

[Image Analyst]

If that still doesn't work, call imopen() or imerode().

[zhifei xu]

Thanks for your answer, with 4-connected, I still got the outer bondary. But what I want is the inner bondary as shown in the right figure.

[zhifei xu]

Great, I succeeded with imopen(),

Thanks a lot

[Image Analyst]

I attached a full demo in my other answer. Thanks for accepting the answer.

Answer 2

[Image Analyst]

Try this:

% Demo by Image Analyst, March, 2021.
clc;    % Clear the command window.
close all;  % Close all figures (except those of imtool.)
clearvars;
workspace;  % Make sure the workspace panel is showing.
format long g;
format compact;
fontSize = 16;
fprintf('Beginning to run %s.m ...\n', mfilename);

%-----------------------------------------------------------------------------------------------------------------------------------
% Read in image.
folder = [];
baseFileName = 'diamond.png';
fullFileName = fullfile(folder, baseFileName);
% Check if file exists.
if ~exist(fullFileName, 'file')
	% The file doesn't exist -- didn't find it there in that folder.
	% Check the entire search path (other folders) for the file by stripping off the folder.
	fullFileNameOnSearchPath = baseFileName; % No path this time.
	if ~exist(fullFileNameOnSearchPath, 'file')
		% Still didn't find it.  Alert user.
		errorMessage = sprintf('Error: %s does not exist in the search path folders.', fullFileName);
		uiwait(warndlg(errorMessage));
		return;
	end
end
grayImage = imread(fullFileName);
[rows, columns, numberOfColorChannels] = size(grayImage)
if numberOfColorChannels == 3
	grayImage = rgb2gray(grayImage);
end
% Display the test image full size.
subplot(2, 3, 1);
imshow(grayImage, []);
axis('on', 'image');
caption = sprintf('Original Gray Scale Image : "%s"', baseFileName);
title(caption, 'FontSize', fontSize, 'Interpreter', 'None');
drawnow;
hp = impixelinfo(); % Set up status line to see values when you mouse over the image.

% Set up figure properties:
% Enlarge figure to full screen.
hFig1 = gcf;
hFig1.Units = 'Normalized';
hFig1.WindowState = 'maximized';
% Get rid of tool bar and pulldown menus that are along top of figure.
% set(gcf, 'Toolbar', 'none', 'Menu', 'none');
% Give a name to the title bar.
hFig1.Name = 'Demo by Image Analyst';

%--------------------------------------------------------------------------------------------------------
% Get histogram
subplot(2, 3, 2);
counts = histcounts(grayImage);
bar(log(counts));
grid on;
title('Log Histogram of gray scale image', 'FontSize', fontSize);
xlabel('Gray Level', 'FontSize', fontSize);
ylabel('Count', 'FontSize', fontSize);

% Threshold image.
mask = grayImage > 128;
% Display mask image.
subplot(2, 3, 3);
imshow(mask, []);
axis('on', 'image');
title('Initial Mask (Binary Image)', 'FontSize', fontSize, 'Interpreter', 'None');
drawnow;
hp = impixelinfo(); % Set up status line to see values when you mouse over the image.

% Label each blob with 8-connectivity, so we can make measurements of it
[labeledImage, numberOfBlobs] = bwlabel(mask, 4);
% Apply a variety of pseudo-colors to the regions.
coloredLabelsImage = label2rgb (labeledImage, 'hsv', 'k', 'shuffle');
% Display the pseudo-colored image.
subplot(2, 3, 4);
imshow(coloredLabelsImage);
hp = impixelinfo(); % Set up status line to see values when you mouse over the image.
caption = sprintf('%d Labeled Blobs', numberOfBlobs);
title(caption, 'FontSize', fontSize, 'Interpreter', 'None');
% Get all the blob properties.  Can only pass in originalImage in version R2008a and later.
blobMeasurements = regionprops(labeledImage, grayImage, 'all');
numberOfBlobs = size(blobMeasurements, 1);

% Do a morphological opening to separate the blobs.
radius = 6;
se = strel('disk', radius, 0);
mask = imerode(mask, se); % Shrink blobs
% Take the largest blob only
mask = bwareafilt(mask, 1);
% Grow them back to the original size.
mask = imdilate(mask, se); % Shrink blobs
% Display mask image.
subplot(2, 3, 5);
imshow(mask, []);
axis('on', 'image');
title('Final Mask', 'FontSize', fontSize, 'Interpreter', 'None');
drawnow;
hp = impixelinfo(); % Set up status line to see values when you mouse over the image.
fprintf('Done running %s.m\n', mfilename);

[zhifei xu]

Thanks a lot! Great works! Does this work for any shaped white regions? I am trying to make it automatically to measure the size of the inner region for different images.

[Image Analyst]

No. It only works if the white region you want is the largest one. Plus you may have to adjust the structuring element size if there is more or less connection between the lagest blob and the connected, smaller ones.

[zhifei xu]

The largest one can be sure. But can I adjust the structuring element size automatically to get what I want? Such as this figure, how can adjust the structuring element size for serval different figures similar to the one in the question and the one below? By the way, I found that the "Threshold of the image" also plays an important role.

Thanks

[Image Analyst]

I think it should work on that. Does it not?

You'll have to come up with some segmentation algorithm. Using a fixed threshold is best but if you can't get control over your image capture conditions, like camera shutter speed or illumination level then you'll have to come up with some sort of automatic threshold selection routine.