uint8 to double is converting pixel values into decimal

when I convert my image which is in uint format to double the pixel matrix values are converting into decimal values

These are uint matrix of image

   181  185  165  128  109  143  163  159  160  164  182  177  179  182  185  187  193  193

and these are double values of same image

0.709803921568628  0.725490196078431  0.647058823529412  0.501960784313726  0.427450980392157  0.560784313725490  0.639215686274510  0.623529411764706  0.627450980392157  0.643137254901961

why is this happening so? whats the solution

Answer 1

[Image Analyst]

Basically it's dividing by 255 when you use im2double(). If you don't want that, then don't do it. If you want the floating point version but the values to be the same, use double() instead of im2double().

[akash bais]

i dont want that to divide by 255, I just want the original integer values of double, how can i get those?

[Guillaume]

Hum, Image Analyst wrote: "If you want the floating point version but the values to be the same, use double() instead of im2double()".

Note that generic non-image processing functions (such as conv2) do not care about the range of values in your image matrix so will work will no problem with double values up to 255. On the other hand, a lot of image processing functions (such as imshow) will not work correctly (or rather as you may expect) if the image type is double but outside the range [0-1].

[Image Analyst]

True. Some work fine with whatever original range of values you're using, like functions imerode(), imdilate(), etc.

Some work with "workarounds", like imshow(dblImage, []) where you can use your original values as long as you use open/close brackets.

Some don't work like you'd expect, like imhist(), unless they're already in the range 0-1 before passing in to the function.

You just have to know what you're doing.

[akash bais]

but my query still remains un answered, or is it so that it isnt possible?

[Image Analyst]

Why is it unanswered? Look:

grayImage = uint8([181 185  165  128  109  143  163  159  160  164  182  177  179  182  185  187  193  193])
dblGrayImage = double(grayImage)
whos dblGrayImage

Look in the command window and you'll see

grayImage =
  1×18 uint8 row vector
   181   185   165   128   109   143   163   159   160   164   182   177   179   182   185   187   193   193
dblGrayImage =
   181   185   165   128   109   143   163   159   160   164   182   177   179   182   185   187   193   193
  Name              Size            Bytes  Class     Attributes

    dblGrayImage      1x18              144  double

You have a double array with the very same values as the uint8 version. I thought that's what you said you wanted? What more do you want?

[akash bais]

okay got it, Thanks a lot sir :)