In this command in "Matlab", what is the purpose of dividing by "256"?>> imshow(filter2(g,c)/256)

Question

In this command in "Matlab", what is the purpose of dividing by "256"?>> imshow(filter2(g,c)/256)

Answer 1

Actually, imshow can handle elements from 0 to 255 if the variable's type is 'uint8'. If the type is double, then it won't scale the colors properly unless they're in the range [0,1]. So rather than dividing by 256, one could also cast the variable to a uint8:

imshow(uint8(filter2(g,c)));

Alternatively, you can use imagesc instead of imshow, which automatically scales its input to maximize the viewable dynamic range. I find the following steps useful for general floating point images:

imagesc(im);

colormap hot;

colorbar;

axis equal;

Hope that helps.

Answer 2

256 is the common number of shades in an 8-bit color channel.

It's likely the output of filter2() is in 8-bit color.

imshow() takes color values in a range from 0 to 1.

Dividing the output of filter2() by 256 converts the color values to something imshow() can work with.

Without knowing what g and c represent though, it's pretty much just a random guess.