MATLAB Answers

0

Take coordinate data from struct, throw it into an array. Why is this hard to find info about?

Asked by Matthew Powell on 16 Feb 2015
Latest activity Edited by Ashish Uthama on 17 Feb 2015
Hi,
I'm working on an image processing program that takes coordinate data of objects in an image acquired via the command regionprops and throws it into an array (that I will grab with a C program).
I'm having no issue sorting singular values into an array, but I am failing to put the coordinate data into one, which looks like [x,y].
code snippet:
blobMeasurements = regionprops(labeledImage, i, 'all'); %labeledImage has object information
numberOfBlobs = size(blobMeasurements, 1);
Centroid = size(blobMeasurements, ???);
g = 1;
n = 1;
for n = 1:97
if (blobMeasurements(n).Area > 500)
Centroid(g) = blobMeasurements(n).Centroid;
g = g + 1;
end
end
All I need is a simple set of lines that can achieve this. For the record, this isn't a homework assignment.

  0 Comments

Sign in to comment.

1 Answer

Answer by Ashish Uthama on 17 Feb 2015
Edited by Ashish Uthama on 17 Feb 2015
 Accepted Answer

Does this help? (If not, use this example code to show how you want your Centriod variable to look like)
stats = regionprops(im2bw(imread('coins.png')),'all');
%%Indices of 'larger' areas
% The key here is using [] to capture the output of stats.Area into an
% array.
inds = [stats.Area]>1000;
%%Centroids, using logical indexing
% This gives you the data in [x1 y1 x2 y2 ....] format
centroids = [stats(inds).Centroid];
[EDIT - add table syntax] With the latest version, regionprops now supports tables, making it even more easier:
stats = regionprops('table',im2bw(imread('coins.png')),'all');
inds = stats.Area>100;
centroids = stats.Centroid(inds,:);

  1 Comment

Yes that works excellently. I also managed to get it to work by right clicking the struct field to create a cell array, and the function was auto-written at the command line. Much appreciated!

Sign in to comment.