Clearing the upper section of a 3-dimensional matrix

2014 2 21
1 个回答
回答已采纳
更新时间：2014 2 22
5 次查看（30 天）

0 个投票

Given a stack of 2D-images, such that they form an x-by-y-by-z volume data set, I would like to delete any data that lies above (column-wise) a set of predetermined column indexes.
For example if:
A(:,:,1) = [1 1 1 A(:,:,2) = [1 1 1 A(:,:,3) = [1 1 1
1 1 1 1 1 1 1 1 1
1 1 1] 1 1 1] 1 1 1]
And the column indexes are defined such as (Where rows correspond to different images):
B = [3 2 1
1 2 3
3 3 3]
The result should be:
C(:,:,1) = [0 0 0 C(:,:,2) = [0 0 0 C(:,:,3) = [0 0 0
0 0 1 1 0 0 0 0 0
0 1 1] 1 1 0] 0 0 0]
My current method involves the use of for-loops, looping through each image and x-position individually, but this proves to be very slow when dealing with large sets of image data. My second thought was to use linear indexing, by converting the column indexes into an array of linear indexes I can easily generate a 3D matrix with 0's at the required column index, but the issue remains of removing the data above those points.
Is there a less computationally intensive method of tackling this problem that I'm missing?
Thanks for any assistance with the matter! :)

3 个评论
显示 1更早的评论 隐藏 1更早的评论

Jan
Jan 2014-2-21
I'm not sure if I see the connection between A, B and C. Please post your code.
Image Analyst
Image Analyst 2014-2-22
Well if B(1,1) = 3, then any elements in A(1,1) above plane 3 should be set to zero. But there are none, since A is only 3 planes high - there are no planes above 3 for A. So C(,1,) should be the same as A for all planes. So C(1,1,:) should equal 1,1,1, which is what A is in the (1,1) column. Yet your C does not show that so I guess I misunderstood or you didn't explain it correctly.
Robert
Robert 2014-2-22
Apologies for the ambiguous example, I meant to illustrate that the masking is being applied up the columns, not between stacks - In fact the images can be treated as separate entities in this example.
A value of B(1,1) = 3 would only imply that C(1:3,1,1) = 0.

请先登录，再进行评论。

请先登录，再回答此问题。

 采纳的回答

David Young
David Young 2014-2-21
编辑：David Young 2014-2-22

0 个投票

Try
Bt = permute(B, [3 2 1]);
mask = bsxfun(@ge, Bt, (1:size(A,1)).');
C = A;
C(mask) = 0;

1 个评论
显示 -1更早的评论 隐藏 -1更早的评论

Robert
Robert 2014-2-22
编辑：Robert 2014-2-22
Thank you! This is much more efficient and makes a huge difference for my application.

请先登录，再进行评论。

更多回答（0 个）

类别

在 帮助中心 和 File Exchange 中查找有关 Matrices and Arrays 的更多信息

标签

Community Treasure Hunt

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

Start Hunting!

Translated by