Every photographer (amateur, enthusiast or pro) knows that getting rid of blurred images is a major issue for all. Thanks to CNet, I recently learned that the camera wobble is now over!
Scientists from the MIT and the University of Toronto worked on the issue and developped a very powerful (as far as it is said) algorithm to automatically remove blur from digital images. Don’t break the news yet! Don’t rush for the champagne! This won’t remove the movements of the photographed subject/model. It’s only targetting the movements of the camera or (rather) the camera holder.
It’s still unsure if this will ever be more efficient than an image stabilizer as we find in many modern digital cameras (digital SLR from Sony or Pentax and lenses from Canon and Nikon, and also more and more digital point-and-shoot compact digital cameras featuring a long zoom lens – that tends to amplify the issue).
But, the real progress, the real leap forward will come when the agorithm will be able to recognize and correct some or all of the movements of the subject itself. This is very easy to understand on the first image illustrating the original CNet article. You will see a young boy playing. Its photo is obviously blurred. But most ofthe problem comes from the fast moves of the young boy not of the photo camera. This is probably why the result appears so limited.