Finding the upper-right "crotches" of a matrix - MATLAB爱好者论坛-LabFans.com

poster · 2019-11-26, 02:00

First, apologies for my description. I don't know the matrix semantics related to this type of operation so I cannot even begin to know how to search for it.

In matlab, I have a matrix A:

   1   1   1   1   1

   1   1   1   1   1

   0   1   1   1   1

   0   0   1   1   1

   0   0   0   0   1

That I would like to transform to become something like B:

   1   0   0   0   0

   0   0   0   0   0

   0   1   0   0   0

   0   0   0   1   0

   0   0   0   0   1

Where the 1's correspond to row-col locations with 1's in the 5 positions above and to the right AND [there is a single 1 either to the left or below OR both left and below positions are zero].

I can see how this might doable by iterating through each position in A, but this may be slow for very large matrices. I was wondering if there is a filter2 or xcorr2 based approach in matlab that might be faster? I tried both with various kernel matrices on a padded version of A (1-padded above and to the right, 0-padded left and below) but can't come to a definitive answer.

2019-11-26, 02:00	#1
poster 高级会员注册日期: 2019-11-21 帖子: 3,006 声望力: 66	Finding the upper-right "crotches" of a matrix First, apologies for my description. I don't know the matrix semantics related to this type of operation so I cannot even begin to know how to search for it. In matlab, I have a matrix A: `1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 0 0 1 1 1 0 0 0 0 1` That I would like to transform to become something like B: `1 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 1 0 0 0 0 0 1` Where the 1's correspond to row-col locations with 1's in the 5 positions above and to the right AND [there is a single 1 either to the left or below OR both left and below positions are zero]. I can see how this might doable by iterating through each position in A, but this may be slow for very large matrices. I was wondering if there is a filter2 or xcorr2 based approach in matlab that might be faster? I tried both with various kernel matrices on a padded version of A (1-padded above and to the right, 0-padded left and below) but can't come to a definitive answer. More answer...