how to detect empty rows and columns from 3-D matrix and crop them?

5 views (last 30 days)
I have a 3-D matrix as a video. It contains empty (0) zero values along the tree dimentions.
The matrix is represented as a video. So I need to:
  • if the row along the frames is empty (==0), I want to crop it.
  • if the column along the frames is empty (==0), I want to crop it
I have an example 3-D dimention but I need a general method.
I have tried to write something but there is something wrong.
Please Help.
%% define the video (3-D matrix)
Frames = [0 0 90 0 0; 0 0 90 0 0; 90 90 90 0 0;0 0 0 0 0;0 0 0 0 0];
Frames(:,:,2 ) = [0 0 0 0 0;0 0 90 0 0; 0 0 90 0 0; 90 90 90 0 0;0 0 0 0 0];
Frames(:,:,3 ) = [0 0 0 0 0;0 0 90 0 0; 0 0 90 0 0; 90 90 90 0 0;0 0 0 0 0];
Frames(:,:,4 ) = [0 0 0 0 0;0 0 90 0 0; 0 0 90 0 0; 90 90 90 0 0;0 0 0 0 0];
Frames(:,:,5 ) = [0 0 0 0 0;0 0 90 0 0; 0 0 90 0 0; 90 90 90 0 0;0 0 0 0 0];
Frames(:,:,6 ) = [0 0 0 0 0;0 0 90 0 0; 0 0 90 0 0; 90 90 90 0 0;0 0 0 0 0];
Frames(:,:,7 ) = [0 0 90 0 0; 0 0 90 0 0; 90 90 90 0 0;0 0 0 0 0;0 0 0 0 0];
Frames(:,:,8 ) = [0 0 0 0 0;0 0 90 0 0; 0 0 90 0 0; 90 90 90 0 0;0 0 0 0 0];
Frames(:,:,9 ) = [0 0 0 0 0;0 0 90 0 0; 0 0 90 0 0; 90 90 90 0 0;0 0 0 0 0];
Frames(:,:,10) = [0 0 0 0 0;0 0 90 0 0; 0 0 90 0 0; 90 90 90 0 0;0 0 0 0 0];
%% define some varibles
%% n & m are zeros counters
%% x,y,z are lengths along the three dimentions
n=0;
m=0;
x=length(Frames(:,1,1));
y=length(Frames(1,:,1));
z=length(Frames(1,1,:));
%% if the row along the frames is empty (==0),crop it.
%% r for row
% %% c for columns
% %% f for frames
for c=1:y
for r=1:x
for f=1:z
if Frames(r,c,f)==0 %detect
m=m+1;
end
end
if m==25
Frames(:,c,:) = []; %crop
end
end
end
%% %% if the column along the frames is empty (==0),crop it.
%% r for row
% %% c for columns
% %% f for frames
for r=1:x
for c=1:y
for f=1:z
if Frames(r,c,f)==0 %detect
n=n+1;
end
end
if m==25
Frames(r,:,:) = []; %crop
end
end
end

Accepted Answer

Matt J
Matt J on 31 Dec 2018
Edited: Matt J on 31 Dec 2018
tmp=any(Frames,3);
I = any(tmp,2);
J= any(tmp,1);
Frames=Frames(I,J,:);

More Answers (2)

Stephan
Stephan on 31 Dec 2018
Edited: Stephan on 31 Dec 2018
Hi,
two lines of code do the job:
Frames = zeros(5,5,10);
Frames(:,:,1 ) = [0 0 90 0 0; 0 0 90 0 0; 90 90 90 0 0;0 0 0 0 0;0 0 0 0 0];
Frames(:,:,2 ) = [0 0 0 0 0;0 0 90 0 0; 0 0 90 0 0; 90 90 90 0 0;0 0 0 0 0];
Frames(:,:,3 ) = [0 0 0 0 0;0 0 90 0 0; 0 0 90 0 0; 90 90 90 0 0;0 0 0 0 0];
Frames(:,:,4 ) = [0 0 0 0 0;0 0 90 0 0; 0 0 90 0 0; 90 90 90 0 0;0 0 0 0 0];
Frames(:,:,5 ) = [0 0 0 0 0;0 0 90 0 0; 0 0 90 0 0; 90 90 90 0 0;0 0 0 0 0];
Frames(:,:,6 ) = [0 0 0 0 0;0 0 90 0 0; 0 0 90 0 0; 90 90 90 0 0;0 0 0 0 0];
Frames(:,:,7 ) = [0 0 90 0 0; 0 0 90 0 0; 90 90 90 0 0;0 0 0 0 0;0 0 0 0 0];
Frames(:,:,8 ) = [0 0 0 0 0;0 0 90 0 0; 0 0 90 0 0; 90 90 90 0 0;0 0 0 0 0];
Frames(:,:,9 ) = [0 0 0 0 0;0 0 90 0 0; 0 0 90 0 0; 90 90 90 0 0;0 0 0 0 0];
Frames(:,:,10) = [0 0 0 0 0;0 0 90 0 0; 0 0 90 0 0; 90 90 90 0 0;0 0 0 0 0];
% crop all empty rows and columns
Frames(:,find(sum(Frames,[1 3])==0),:)=[];
Frames(find(sum(Frames,[2 3])==0),:,:)=[]
disp('Please accept useful answers.')
Best regards
Stephan
  2 Comments
Nibras Abo Alzahab
Nibras Abo Alzahab on 31 Dec 2018
Thank you for your answer.
I got an error while using the
Error using sum
Dimension argument must be a positive integer scalar within indexing range.
Frames(:,find(sum(Frames,[1 3])==0),:)=[];
Stephan
Stephan on 31 Dec 2018
Edited: Stephan on 31 Dec 2018
Hm, works for me. Which Release do you use? however - shortest answer from Matt is a nice one.

Sign in to comment.


Luna
Luna on 31 Dec 2018
Edited: Luna on 31 Dec 2018
Hi,
Try this below:
%% delete zero rows
Frames = reshape(Frames(bsxfun(@times,any(Frames,2),ones(1,size(Frames,2)))>0),[],size(Frames,2),size(Frames,3));
%% delete zero columns
Frames = reshape(Frames(bsxfun(@times,any(Frames,1),ones(size(Frames,1),1))>0),size(Frames,1),[],size(Frames,3));
you will get 3x3x10 matrix
  6 Comments
Nibras Abo Alzahab
Nibras Abo Alzahab on 31 Dec 2018
Thank you for your intrest and ansewrs.
Luna is right
I souhld get 4x3x10 matrix.
But in your answer: 3x3x10 matrix

Sign in to comment.

Products


Release

R2018b

Community Treasure Hunt

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

Start Hunting!