imwrite and cdata as double

[bpabbott]

I encountered an error using imwrite today.

When the cdata is of type double it is converted by ...

img = uint8 (img - 1);

This produces an error because the lowest color index is 1.

I'm unfamiliar with images and imwrite so I thought it prudent to check before pushing the attached change.

Ben

changeset.patch
Description: Binary data