new_img = convert(img, text);
(img, text) = convert_back(new_img);
Can someone illustrate with a built-in image of MATALB?
new_img = convert(img, text);
(img, text) = convert_back(new_img);
Can someone illustrate with a built-in image of MATALB?
I believe you're looking for steganography. You can start with this MATLAB implementation of LSB steganography.
A simple way of doing LSB steganography is to take an lossless-compressed image and set the LSB of each component (R,G,B). Then for a m x n image you get 3mn bits to store information in. Since you're modifying the LSB, the difference will not be perceivable in the image.
Update
SO I decided to code up a small, inefficient but demonstrative example:
function LSBStega
%%// Image and text
I = imread('coins.png');
text = 'Hello World etc';
assert(numel(I) > numel(text)*8,'Insufficient number of pixels');
%%// Encode
%// The end character signifies the end of the hidden text
end_char = '.';
%// Append it
text = [text end_char];
%// Convert each character into binary form with atleast 8 bits
%// We transpose it before calling (:) since the binary representation
%// is stores each character in binary per row and the (:) operations
%// vectorizes the matrix by column.
b = transpose(dec2bin(text,8));
%// Find which bits should be set to 1 or 0
ind_0 = find(b == '0');
ind_1 = find(b == '1');
%// Set the appropriate bits to 1 and 0 and leave the rest alone
I(ind_0) = bitset(I(ind_0),1,0);
I(ind_1) = bitset(I(ind_1),1,1);
%%// Faster decode
text_back = [];
for i = 1:8:numel(I)
%// Extract the LSB from a set of 8 bytes in the image
C = bitget(I(i:i+7),1);
%// Convert from binary to decimal
C = bin2dec(num2str(C));
%// Check if it's the end character; break if so, store if not
if(C == end_char)
break;
else
text_back(end+1) = C;
end
end
%// Convert to text
text_back = char(text_back);
%%// Display
subplot(1,2,1);
title('Original');
imshow(imread('coins.png'));
subplot(1,2,2);
title('Steganography Result');
imshow(I);
disp(text_back);
end