It could also be used to detect copyright infringement. If google could “reverse search” an image I created, I would know who was using or storing my image without my permission.

However, I wonder how this would affect generated pages. That is, if I was searching a header file that was on every page of a php-based site, which pages would be relevant to my initial search? What about applying the boolean search rules to this for a more complex reverse search?

Oh! I was going to suggest something that would combine this idea with Dr. Wang’s ALIPR project. That is, submit an image and similar images are displayed or perhaps related pages. But, it appears this feature is already a part of Dr. Wang’s project. Cool stuff.