The 'porthole' approach
With this approach I would place a small rectangle on any area of a random Google image and, with a gun-to-head mentality, force an idea from the selected area. It can help to take a step or two away from the monitor.
I think the area in the 'porthole' has some visual appeal.
Images within images
With this approach I would peruse pages of random Google images and pick out the images that contain images within themselves and focus on those images.
I could also use this approach as a general creativity tool. As with peripheral viewing, I would set a directive such as:
What do you see in this image?
With the image above I initially thought it looked like a dog with a barrel beneath its neck and then I thought it looked like George W Bush with an inferno behind him.
I like the image-within-image approach because of the ambiguity of the images. An opportunity to let the imagination run riot :)