Watermarking
by Jonas Öberg. Average Reading Time: about 2 minutes.
Here’s an image of yours truly, taken by Creative Commons’ own David Kindler. It’s an example of a watermark in an image, this time to announce that the picture is from the Global Summit 2011, but which could equally be used to say that this picture was taken by David Kindler, as a way of ensuring that the picture was correctly attributed when it was re-used by me in this post.
CC Global Summit-Jonas Oberg.jpg | Flickr – Photo Sharing!DTKindler Photohttp://creativecommons.org/licenses/by/2.0/deed.en / CC BY 2.0
The problem with this kind of watermarking is partly that it takes away something from the image: it’s an invasive procedure that modifies the content of the image. Even if we were to resize the canvas of the picture so that the addition of the watermark fit outside of the actual picture, or perhaps to figure out a way to do minimalistic changes of individual pixels to somehow engrain the information without the picture, it only partially work.
If I were to rotate the picture 90 degrees for publishing, the attribution ends up on the side. Since this is licensed under a Creative Commons Attribution license, I might also chose to take just part of the image, just ignoring the watermark all together. And if I resize the picture, the attribution might get so small that it just can’t be read.
It’s a solution, but it’s not an ideal one. That’s why we’re working with the concept of metadata: information which is recorded as part of the image, but not in the visual representation of it. EXIF is the most common form of metadata for images, it allows a camera to record information about what aperture was used when shooting the image, if the flash was used or not, sometimes where the image was taken, and a lot of other information. Data which is useful to have, but which should not be part of the image itself.
Adding information about licensing and creator to this metadata would allow us to create tools that read this metadata (or a link to the metadata) and understand what to do with it, such as automatically crediting the creator when we make use of the image, or, if we try to use the image in a way that the author doesn’t want, we could get a helpful hint from our software letting us know that we might want to look into our use of the image. It shouldn’t prevent us, but it should give us notice.
read original post on [author-name]'s Site