how to reverse a random permutation of a matrix to get back the original matrix?

10 visualizaciones (últimos 30 días)
O = key1(randperm(numel(key1)));
here O is the random permutation of rows and columns of a 'key1' matrix which is '16*147'. how to get back my original key1 matrix?

Respuesta aceptada

Andrei Bobrov
Andrei Bobrov el 17 de Sept. de 2018
key1 = randi(398,4,5);
s = size(key1);
p = reshape(randperm(numel(key1)),s);
O = key1(p);
[~,ii] = sort(p(:));
out = O(reshape(ii,s));
  1 comentario
Bruno Luong
Bruno Luong el 17 de Sept. de 2018
Editada: Bruno Luong el 17 de Sept. de 2018
alternatively one can do indexing on the lhs and avoid sorting
out(p) = O;
out = reshape(out,s);

Iniciar sesión para comentar.

Más respuestas (1)

Image Analyst
Image Analyst el 17 de Sept. de 2018
See attached demo.
  2 comentarios
shashwat soni
shashwat soni el 17 de Sept. de 2018
in my case the image(original) matrix is a 2d matrix, now i want to generate 3 random integers for each element of a row in the matrix, for e.g. if my matrix is 16*49, so i want a 16*147(147 = 49*3) matrix so that each element of a row in the original matrix gives me 3 random elements. And again i should be able to decrypt it
Image Analyst
Image Analyst el 17 de Sept. de 2018
My code would give 16*49 numbers, which would keep each pixel together. In other words, the green and blue values move to the same location as the red value. This has the effect of simply shuffling the pixels to new locations while keeping the color of the pixel the same (because red, green, and blue all move together to the same new location).
I'm not sure if you want 3 values for each row, or if you want 3 values for each pixel - you seemed to say/indicate both. If you wanted 3 values for each row, I'm not sure how that would look. If you wanted 3 values for each pixel then you could be moving the red, green, and blue values to different locations. To do that with my code you'd simply get 3 sort orders, one for each color channel:
% Get the orders to scramble them in.
scrambleOrderR = randperm(rows*columns);
scrambleOrderG = randperm(rows*columns);
scrambleOrderB = randperm(rows*columns);
% Scramble according to the scrambling order for each individual color channel.
redChannel = redChannel(scrambleOrderR);
greenChannel = greenChannel(scrambleOrderG);
blueChannel = blueChannel(scrambleOrderB);

Iniciar sesión para comentar.

Categorías

Más información sobre Creating and Concatenating Matrices en Help Center y File Exchange.

Community Treasure Hunt

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

Start Hunting!

Translated by