How can I extract a portion of an image specified by mask generated from a ROI in Image Processing Toolbox 7.3 (R2011b)?

56 vues (au cours des 30 derniers jours)
I have an image where I have specified a region of interest (ROI) using functions such as IMELLIPSE or IMFREEHAND. I would like to extract the part of the image contained within the region specified by such a function.

Réponse acceptée

MathWorks Support Team
MathWorks Support Team le 14 Fév 2012
We first create a binary mask from a specified ROI, which returns a matrix containing only 1 or 0. To return the selected portion of the image, we perform a logical AND of the binary mask matrix with the image using logical indexing. The resulting image will contain the extracted portion.
The steps below provide an example of this process. Here, we use a sample image provided by the Image Processing Toolbox.
% 1) Open the image file as a workspace variable:
img = imread('peppers.png');
% 2) Display the image and store its handle:
h_im = imshow(img);
% 3) Create an ellipse defining a ROI:
e = imellipse(gca,[55 10 120 120]);
% 4) Create a mask from the ellipse:
BW = createMask(e,h_im);
% 4a) (For color images only) Augment the mask to three channels:
BW(:,:,2) = BW;
BW(:,:,3) = BW(:,:,1);
% 5) Use logical indexing to set area outside of ROI to zero:
ROI = img;
ROI(BW == 0) = 0;
% 6) Display extracted portion:
figure, imshow(ROI);
  1 commentaire
Image Analyst
Image Analyst le 3 Juil 2018
Yes. Just use BW as a logical index for img (or ROI, the copy of img):
img(BW) = 0; % Set img = 0 inside the BW mask.
See attached demos to learn more.

Connectez-vous pour commenter.

Plus de réponses (1)

michael scheinfeild
michael scheinfeild le 26 Fév 2019
if you have some mask (nr*nc image)use find where the indexes are 1 in the mask only for rectangular mask !
okind=find(Mask>0);
[ii,jj]=ind2sub(size(Mask),okind);
ymin=min(ii);ymax=max(ii);xmin=min(jj);xmax=max(jj);
imCropped=imcrop(Y,[xmin,ymin,xmax-xmin+1,ymax-ymin+1]);

Produits


Version

R2011b

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by