Not (~) operation of binary image after subtracting double data type of binary images. Could I have the explaination of this?

I have 2 binary images. I substracted it, then I did logical operation of not (~)

im1=imread('im1.jpeg');
im2=imread('im2.jpeg');
im3=(double(im1)-double(im2));
im4=~im3;

This is the image after subtracting the binary images which are not double data type.

After subtracting the images, I do logical operation of not (~).

im3=im1-im2;%without double data type
im4=~im3;

This is the result without double data type

Please, give me an explaination. Thank you!

Answer 1

[Matt J]

If im1 and im2 start off as uint8, then subtraction operations 0-1 will yield zero values,

>> uint8(0)-uint8(1)

ans =

  uint8

   0

unlike when they are doubles.

[Aaron Abel]

Thanks, I have analyzed. There are -255 intensity of the pixels that is changed to 0 by not operation.