How shall I ignore the reflection in an image and select the rest of the image in a bounding box and obtain the results also automate it in order to process multiple images?

I have a number of images of a combined droplet spreading on a surface. I need to do image processing in order to calculate the droplet spreading length and height as indicated in the attached images. I have run a code which binarizes and thresholds the image and a applies a boundingbox to capture the dimension of the blob. Now it can be seen from the image that there is a reflection of the combined droplet blob which is visible as the solid substrate is glass. The Matlab code accurately captures the spreading width horizontally through the bounding box but for the vertical direction it includes the reflected portion as well. So how can it be done? Also how could I automate it so the code loads a series of images (say 25 images) processes those image files and returns both the pixel size of the length and breadth?

What modifications on this code could be incorporated

clc;    % Clear the command window.
close all;  % Close all figures (except those of imtool.)
clear;  % Erase all existing variables. Or clearvars if you want.
workspace;  % Make sure the workspace panel is showing.
format long g;
format compact;
fontSize = 18;
folder = 'C:\Users\Admin\Desktop\Droplet images for calculation\0.50 cmc on water\0.5 cmc head on\fixing time'; 
baseFileName = '0.5 cmc head on0211.tif';
fullFileName = fullfile(folder, baseFileName);
% Check if file exists.
if ~isfile(fullFileName)
	% 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
	fullFileName = fullFileNameOnSearchPath;
end
[grayImage,map] = imread(fullFileName);
grayImage = rgb2gray(grayImage);
imshow(grayImage,map)
binaryImage = ~imbinarize(grayImage);

% Erase from line 758 down:
binaryImage(687:end, :) = false; % 758

% Fill holes.
binaryImage = imfill(binaryImage, 'holes');

% Get rid of any msall noise blobs.
binaryImage = bwareafilt(binaryImage, 1); % Take largest blob only.
imshow(binaryImage)
impixelinfo;

props = regionprops(binaryImage, 'BoundingBox');
spreadingWidth = props.BoundingBox(3)
droplet_height = props.BoundingBox(4)
rectangle('Position', props.BoundingBox, 'Edgecolor', 'g', 'LineWidth', 2)
fprintf('Done running %s.m ...\n', mfilename);

Answer 1

[Image Analyst]

Rather than remove it, just scan across your binary image finding where it's white

[rows, columns] = size(binaryImage)
topRows = nan(1, columns);
for col = 1 : columns
    t = find(binaryImage(:, col), 1, 'first');
    if ~isempty(t)
        topRows(col) = t;
    end
end
inDroplet = ~isnan(topRows);
leftColumn = find(inDroplet, 1, 'first')
rightColumn = find(inDroplet, 1, 'last')
spreadingWidth = rightColumn - leftColumn; % Pixel center to pixel center.  Add 1 if you want whole pixels.

[Pragyan Kumar Sarma]

What exactly the above code will implement, could you please tell a bit about it? I was trying again to erase from the line 681

% Erase from line 758 down:
binaryImage(681:end, :) = false; % 758

This eliminates the reflection part which I feel now works for me.

[Image Analyst]

It scans across looking at your binary image for a white blob. If there is no white blob, the topRows value will still be nan but columns where it's white will have the row where the white is. So, because your reflection is totally underneath the stuff on top, we can use my code to determine the maximum horizontal extent of the blob, which is the width you want.

The problem with your code is that is depends on the row being exactly 681 (or 758) and if it's not that, then you won't be erasing the right part of the image. My code has no such limitation and is more robust and adaptable to possible changes in your image. If you can find the width, like I do, then there is no reason to erase down to the bottom of the image - it's just not necessary.

Also, to automate it, see the code in the FAQ:

Process a sequence of files

[Pragyan Kumar Sarma]

Thanks for your answer. I also need to get the vertical distance of the coalesced drop, so in order to do so I need to ignore the reflected part of the image. So any more modification that could scan and identify the vertical distance part?

For this does it need to be something like

spreadingHeight = topRow - bottomRow

[Image Analyst]

You can do

spreadingHeight = range(topRow)

or

spreadingHeight = max(topRow) - min(topRow)