Can I represent an image in a binary tree format?

3 次查看（过去 30 天）
srikanth Appala
srikanth Appala 2013-3-19
if true
% code
end

请先登录，再回答此问题。

回答（2 个）

Image Analyst
Image Analyst 2013-3-19
See qtdecomp() in the Image Processing Toolbox.
  6 个评论
显示 4更早的评论
srikanth Appala
srikanth Appala 2013-3-21
sir, my project is "Binary partitioning tree as an efficient representation for image segmentation, information retrieval and image processing".so binary partitioning with image regions is important ro to us.to devide image into regions "region merging algorithm" is used,bt i dont knw how construct tree.help me
Image Analyst
Image Analyst 2013-3-21
I've used qtdecomp only briefly once and that was to just understand how it works. I never need to do that. It doesn't return some information that I needed and when I called the Mathworks they weren't too clear on how it worked either. Anyway, I don't use it. Not sure why you think you need to do this or why you chose that project subject. Can you explain why? Better yet, start your own discussion, rather than intertwine your discussion with Allesandro's.

请先登录，再进行评论。

Alessandro
Alessandro 2013-3-21
编辑：Alessandro 2013-3-21
from wikipedia I read the following about image segmentation:
In computer vision, image segmentation is the process of partitioning a digital image into multiple segments (sets of pixels, also known as superpixels). The goal of segmentation is to simplify and/or change the representation of an image into something that is more meaningful and easier to analyze
You need a tree and the "superpixels" values of the tree. I just wannted to understand the sparse objects from matlab so I tryed the qtdecomp function:
%define some grayscale image
I = uint8([1 1 1 1 2 3 6 6;...
1 1 2 1 4 5 6 8;...
1 1 1 1 7 7 7 7;...
1 1 1 1 6 6 5 5;...
20 22 20 22 1 2 3 4;...
20 22 22 20 5 4 7 8;...
20 22 20 20 9 12 40 12;...
20 22 20 20 13 14 15 16]);
%Get where there is information
S = qtdecomp(I,.05);
%Get the information using the simply mean value
erg = sparse(0);
blocks = unique(nonzeros(S));
for blocksize = blocks'
[y x] = find(S==blocksize);
for i=1:length(x)
erg(x(i),y(i)) = mean2(I(y(i):y(i)+blocksize-1,x(i):x(i)+blocksize-1));
end
end
rebuildimage = zeros(size(S));
%Rebuild the image from the mean values in the block
for blocksize = blocks'
[y x] = find(S==blocksize);
for i=1:length(x)
rebuildimage(y(i):y(i)+blocksize-1,x(i):x(i)+blocksize-1) = nonzeros(erg(x(i),y(i)))
end
end
disp(rebuildimage)
So now you can see rebuildimage looks like I. In the matlab sparse arrays S and erg you have the "super pixels" information.
  4 个评论
显示 2更早的评论
Mohammed
Mohammed 2013-4-4
编辑：Mohammed 2013-4-4
It is useful for things like topology representation of the segmentation maps. It saves a lot of time for searching algorithms instead of doing linear searching, they use them as a probabilistic framework of searching that can reduce the time by a huge factor (from week to 10 mins of runtime)... I would suggest you to read about huffman coding and binay trees for more understanding about the tree representation!

请先登录，再进行评论。

类别

在 Help Center 和 File Exchange 中查找有关 Convert Image Type 的更多信息

标签

产品

Community Treasure Hunt

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

Start Hunting!

Translated by