To RESHAPE the number of elements must not change?

here is the full code of the program:

%%Main Function
function ImageSharing
clear; close all;
Height = 256;
Width =256;
t = 4;
w = 8;
Users = [1,3,5,8];
base = 257;

Sub_width  = Width/t^(0.5);
Sub_height = Height/t^(0.5);

if(length(Users)<t)
    error('invalid length of Users');
end;

if(mod(Sub_height,1) ~= 0)
    error('invalid length of Sub height');
end;

Im_Name = 'i.bin';
[Im_linear, Im_Square] = ReadImage(Im_Name,Height,Width);
P_Im = GetPreparedIm(Im_linear,t);

Shadow_Image = GetShadow_Image(P_Im,w,base);

a = find(Shadow_Image>256);
ShowShadowImage(Shadow_Image,Sub_width,Sub_height);

RequiredShadowImage = GetRequiredShadowImage(Shadow_Image,Users);

% RequiredShadowImage = AddNoise(RequiredShadowImage);

RecoveredImage = RecoverImage(RequiredShadowImage,Users,Height,Width,base);

Err = sum(RecoveredImage(:) - Im_linear);
disp(sprintf('Error bit: %d',Err));
figure;
imshow(uint8(RecoveredImage));
title('Recovered Image');

%%function RequiredShadowImage = AddNoise(RequiredShadowImage);
function RequiredShadowImage = AddNoise(RequiredShadowImage)
TT = size(RequiredShadowImage);
Err = round(rand(TT(1),TT(2)));
Err(Err == 0) = -1;
RequiredShadowImage = RequiredShadowImage + int32(Err); 

%%function RecoveredImage = RecoverImage(RequiredShadowImage)
function RecoveredImage = RecoverImage(RequiredShadowImage,Users,Height,Width,base)
TT = size(RequiredShadowImage);
xx = ones(TT(2),TT(2));
for i = 1:TT(2)
   xx(:,i) = xx(:,i).*(Users.^(i-1))';
end;
for i = 1:TT(1)
    if(mod(i,100)==0)
        disp(sprintf('%d of %d',i/100,floor(TT(1)/100)));
    end;
    RecoveredImage(i,:) = solveEq(xx,RequiredShadowImage(i,:),base);
end;
RecoveredImage = reshape(RecoveredImage(:),Height,Width)';

%%function b = solveEq(Users,RequiredShadowImage(i,:),TT(2),base);
function b = solveEq(xx,yy,base)
b = GetIntMod(inv(sym(xx))*yy',base)';
%b = inv(sym(xx))*yy';

%%function Re_b = GetIntMod(b,base)
function Re_b = GetIntMod(b,base)
[n,d] = numden(b);
n = double(n);
d = double(d);
Re_b = mod(n.*powermod(d,-1,base),base);

%%function RequiredShadowImage = GetRequiredShadowImage();
function RequiredShadowImage = GetRequiredShadowImage(Shadow_Image,Users)
TT = length(Users);
for i = 1:TT
    RequiredShadowImage(:,i) = Shadow_Image(:,Users(i));  
end;

%%ShowShadowImage
function ShowShadowImage(Shadow_Image,Sub_width,Sub_height)
TT = size(Shadow_Image);
for i =1:TT(2)
    figure;
    imshow(uint8(reshape(Shadow_Image(:,i),Sub_width,Sub_height)));
    cmd = sprintf('title(''Shadow Image %d\'')',i);
    eval(cmd);
end;

%%get Sharing_Polynomial
function Shadow_Image = GetShadow_Image(P_Im,w,base)
TT = int32(size(P_Im));
P_Im = int32(P_Im);
w=int32(w);
base = int32(base);
Shadow_Image = int32(zeros(TT(1),w)); 
for i=1:w
    for j = 1:TT(2)
        Shadow_Image(:,i) = Shadow_Image(:,i) + mod(P_Im(:,j)*i^(j-1), base);
    end;
end;

   Shadow_Image = mod(Shadow_Image,base);

%%get PreparedIm
function P_Im = GetPreparedIm(Im_linear,t)
if(mod(length(Im_linear),t) ~= 0)
    error('invalid length')
end;
P_Im = reshape(Im_linear,[],t);

%%Read Image
function [Im_data, Im]=ReadImage(Im_Name,Height,Width)
Im_fp = fopen(Im_Name,'rb');
Im_data = fread(Im_fp,'uint8');
h = figure;   % create a new figure
Im = uint8(reshape(Im_data,Height,Width));
imshow(Im);
title('Original Image');

function y = powermod(a,z,n)
% This function calculates y = a^z mod n
% If a is a matrix, it calculates a(j,k)^z mod for every element in a
[ax,ay]=size(a);

% If a is negative, put it back to between 0 and n-1
a=mod(a,n);

% Take care of any cases where the exponent is negative
if (z<0),
   z=-z;
   for j=1:ax,
      for k=1:ay,
         a(j,k)=invmodn(a(j,k),n);
      end;
   end;   
end;

for j=1:ax,
for k=1:ay,
  x=1;
  a1=a(j,k);
  z1=z;
  while (z1 ~= 0),
     while (mod(z1,2) ==0),
        z1=(z1/2);
        a1=mod((a1*a1), n);
     end;  %end while
     z1=z1-1;
     x=x*a1;
     x=mod(x,n);
  end;
  y(j,k)=x;  
end; %end for k
end; %end for j

function y = invmodn( b,n);
% This function calculates the inverse of an element b mod n
% It uses the extended euclidean algorithm

n0=n;
b0=b;
t0=0;
t=1;

q=floor(n0/b0);
r=n0-q*b0;
while r>0,
   temp=t0-q*t;
   if (temp >=0),
      temp=mod(temp,n);
   end;
   if (temp < 0),
      temp= n - ( mod(-temp,n));
   end;
   t0=t;
   t=temp;
   n0=b0;
   b0=r;
   q=floor(n0/b0);
   r=n0-q*b0;
end;

if b0 ~=1,
   y=[];
   disp('No inverse');
else
   y=mod(t,n);
end;   
Here is the error:
??? Error using ==> reshape
To RESHAPE the number of elements must not change.

Error in ==> ImageSharing>ReadImage at 127 Im = uint8(reshape(Im_data,Height,Width));

Error in ==> ImageSharing at 24 [Im_linear, Im_Square] = ReadImage(Im_Name,Height,Width); Whenever i run this with an image i get the error in the reshape statement. Please help.

[Stephen23]

@ Imdad Choudhury: please edit your question and give us the complete error message. This means all of the red text. It has information that tells us what the code was doing. Otherwise we have to use our crystal balls...

[Imdad Choudhury]

Here it is.

??? Error using ==> reshape To RESHAPE the number of elements must not change.

Error in ==> ImageSharing>ReadImage at 127 Im = uint8(reshape(Im_data,Height,Width));

Error in ==> ImageSharing at 24 [Im_linear, Im_Square] = ReadImage(Im_Name,Height,Width);

[Stephen23]

Perhaps there might be some confusion between reshape and resizing the image... note that reshape does not change how many array elements there are (you can think of this as the number of pixels). If you want to change how many pixels your image has (e.g. to make any image have 255x255 pixels), then reshape is the wrong tool for you: correct would be imresize.

Answer 1

[dpb]

All the code is superfluous and a red herring to wade through excepting for the function containing the error..

function [Im_data, Im]=ReadImage(Im_Name,Height,Width)
Im_fp = fopen(Im_Name,'rb');
Im_data = fread(Im_fp,'uint8');
h = figure;   % create a new figure
Im = uint8(reshape(Im_data,Height,Width));

You've passed in an x- and y-dimension which together define the precise number of elements but you read a file which can be any dimension and then try to reshape it to that exact size. If they're not the same, this is going to fail (which it did in your case).

Use debugger and step through to see what

size(Im_data)

returns, unfortunately you're going to find that numel(Im_data) is not going to be Height*Width

I can't tell you what you want to do to compensate for the difference in sizes, but that's your error.

[Imdad Choudhury]

@dpb Can you please simplify and tell me what to do. ? That would be greatly helpful. Ans also both size and numel returns 3087.

[dpb]

What're Height and Width?

[Imdad Choudhury]

I have taken them so the images are of fixed size. This code works fine when I take the 'lena.bin' file. but when I take other images it doesn't work.

[Walter Roberson]

Most images are RGB images which are THREE dimensional. Your code expects them to be TWO dimensional.

[Walter Roberson]

When you do not tell fread() how many values to read, it reads all of the values in the file. That might not be the same as the supplied height and width, especially if the file contains information for R G and B.

You can patch around the problem by telling fread() how much data to read:

Im_data = fread(Im_fp, [Height, Width], 'uint8');

and you will not even need the reshape() afterwards. However, if you use this and the file contains RGB data, you are only going to end up reading 1/3 of the available data.

[Imdad Choudhury]

@Walter Roberson- I have taken only grayscale images for that purpose.

[Walter Roberson]

After the fread() show us size(Im_data) and show us Height and Width

[Imdad Choudhury]

K>> Height

Height =

   256

K>> size(Im_data)

ans =

      185022           1

K>> Width

Width =

   256

[Walter Roberson]

That file does not contain (only) uint8 data for a 256 x 256 matrix. The size of the file, 185022, could represent uint8 data for 342 x 541, or it could be that the file is an image file that represents compressed data.

Please show us code that you used to create the file.

[Imdad Choudhury]

I haven't created the images in different file. I am using some of the default images that are used generally for this kind of work.

[Walter Roberson]

Where did you get the other image? How do you know that they are uint8 with no header and 256 x 256 ?

[Imdad Choudhury]

I downloaded the images. Now I understand where the problem lies. So is there any way to use the reshape instruction in some dynamic way and also by not fixing the height and width ?

[Stephen23]

Perhaps there might be some confusion between reshape and resizing the image... note that reshape does not change how many array elements there are (you can think of this as the number of pixels). If you want to change how many pixels your image has (e.g. to make any image have 255x255 pixels), then reshape is the wrong tool for you: correct would be imresize.

[Walter Roberson]

You need to know either the height or the width for the binary image (or both). If you know both then you can use

Im_data = fread(Im_fp, [Height, Width], 'uint8');

with no reshape() needed.

If you know the height but not the width then you can use

Im_data = fread(Im_fp, [Height, inf], 'uint8');

with no reshape() needed.

If you know the width but not the height then,

Im_data = reshape( fread(Im_fp, 'uint8'), [], Width);

If you do not know either the height or the width then you can cannot be sure what the size of the image is. For example that 185022 could be 342 x 541 or could be 541 x 342 or could be 171 x 1082 or some other combinations of sizes.

[Imdad Choudhury]

i tried all of these but it didn't solve the problem. the error then shifts to the other reshape lines.. i got the code from here http://in.mathworks.com/matlabcentral/fileexchange/39660-secret-image-sharing-scheme The author isn't available so i had to ask here.

[aisha khan]

I also need help with this code...i am stuck at the last part..what does the function powermod and invmodn do..is it related to overflow prevention?