'dec2bin' produce error when converting rounded float number

I tried to load an image, scale and output the binary code of each pixel. 'dec2bin' requires integer smaller than 2^52, so I rounded each floating element and then did the conversion, however the error keeps jumping out saying: ----------------------------------------------

*??? Error using ==> dec2bin at 33

D must be a non-negative integer smaller than 2^52.

Error in ==> image_ge_v2 at 12

v = dec2bin(iresize_r(i),9);*

----------------------------------------------

It is for sure that the elements in the matrix are smaller than 256.

Here's the segment of my code:

----------------------------------------------

icolour = imread('iig_v22.jpg');

icolour_s = imresize(icolour,[200 200]);

igray = rgb2gray(icolour_s);

igray = double(igray);

A = fopen('image.txt','wt');

iresize = imresize(igray,[60 60]);

iresize_r = round(iresize);

for i = 1:numel(iresize_r)

v = dec2bin(iresize_r(i),9);

end

------------------------------------------------

Please give me some hint!

Thanks a lot

Answer 1

[Sean de Wolski]

icolour = imread('peppers.png');
icolour_s = imresize(icolour,[200 200]);
igray = rgb2gray(icolour_s);
igray = double(igray);
iresize = imresize(igray,[60 60]);
iresize_r = round(iresize);
for i = 1:numel(iresize_r)
   v = dec2bin(iresize_r(i),9);
end

I get no error doing this with a stock image.

[Jeff]

well the trick is here, I also tried different image some of them are fine but others produce error !

Answer 2

[Fangjun Jiang]

We don't have your data so probably won't re-produce your error. You could check the value of i when the error happened and then check the value of iresize_r(i).

Once your problem is solved, you can use dec2bin(iresize_r,9) without the for-loop.