Calculate the measurements of a Binary Image with an ndims of 3

I am analyzing a series of images that increase the volume of the yellow plume over a series of time steps. My goal is to compare the area, centroid, outline and lengths of the largest finger of each yellow plume with other similiar images in a table as well as have MATLAB show the centroid and extrema on the image. In order to measure these properties, Matlab states I need a ndims of 2 or less where my binary image is 3. I used the imbinarize command to create a binary image and I recieved this error using regionprops.

An example of the photo to be analyzed is attached to this post.

Thanks for your help.

Ryan

[Rik]

Did you convert the RGB image to grayscale before binarizing?

[Ryan Tigera]

Hi Rik,

Thanks for the help. In reviewing my code, I skipped this step.

Thanks again for your help.

Answer 1

[Shashank Gupta]

Hey Ryan,

you might need to convert your RGB image to Gray image first and then binarize it, this will make sure you have ndim equals to 2. Attaching a small piece of code for your referece.

% Read image
img = imread('A,T=8min,V=40mL,TR.JPG');
% Convert RGB to gray.
im = rgb2gray(img);

% Binarize.
im = imbinarize(im);
% Use regionprop function according to your necessity.
stats = regionprops('table',im,'Centroid',...
'MajorAxisLength','MinorAxisLength')

Hope this helped.

Cheers.

[Rik]

Just a minor correction: rgb2gray will reduce ndims to 2. Even scalars have 2 dimensions in Matlab. rgb2gray reduces the size of the third dimension to 1, which is probably the source of this confusion.

[Shashank Gupta]

Yes rik, my mistake, I will edit it and correct it. Thank for noticing.

[Ryan Tigera]

Thanks to you both. This has resolved my issue